Anchor Box Optimization for Object Detection

Size: px
Start display at page:

Download "Anchor Box Optimization for Object Detection"

Transcription

1 Anchor Box Optmzaton for Object Detecton Yuany Zhong 1, Janfeng Wang 2, Jan Peng 1, and Le Zhang 2 1 Unversty of Illnos at Urbana-Champagn 2 Mcrosoft Research 1 {yuanyz2, 2 {janfw, arxv: v1 [cs.cv] 2 Dec 2018 Abstract In ths paper, we propose a general approach to optmze anchor boxes for object detecton. Nowadays, anchor boxes are wdely adopted n state-of-the-art detecton framewors. However, all these framewors pre-defne anchor box shapes n a heurstc way and fx the sze durng tranng. To mprove the accuracy and reduce the effort to desgn the anchor boxes, we propose to dynamcally learn the shapes, whch allows the anchors to automatcally adapt to the data dstrbuton and the networ learnng capablty. The learnng approach can be easly mplemented n the stochastc gradent descent way and be plugged nto any anchor box-based detecton framewor. The extra tranng cost s almost neglgble and t has no mpact on the nference tme cost. Exhaustve experments also demonstrate that the proposed anchor optmzaton method consstently acheves sgnfcant mprovement ( 1% map absolute gan) over the baselne method on several benchmar datasets ncludng Pascal VOC 07+12, MS COCO and Branwash. Meanwhle, the robustness s also verfed towards dfferent anchor box ntalzaton methods, whch greatly smplfes the problem of anchor box desgn. 1. Introducton Object detecton plays an mportant role n many real applcatons and recent years have seen great mprovement n terms of speed and accuracy based on neural networs [18, 16, 17, 13, 11]. Many of these modern deep learnng based detectors mae use of the anchor boxes (or default boxes), whch serves as the ntal guess of the boundng box. These anchor boxes are densely dstrbuted across the output feature map, typcally centered at each neuron of the feature map. The neural networ s traned to predct the poston offset relatve to the cell center (sometmes normalzed by the anchor sze) and the wdth/heght offsets relatve to the anchor box shape, as well as the classfcaton confdence. One of the crtcal factors s the desgn of the anchor wdth and the anchor heght, and most of the approaches determne the values by ad-hoc heurstc methods. For nstance of Faster R-CNN[18], the anchor shapes are of 3 scales (128 2, 256 2, ) and of 3 aspect ratos (1 : 1, 1 : 2, 2 : 1). In SSD[13], the aspect ratos also nclude 1 : 3 and 3 : 1 wth multple scales for dfferent feature maps. The approach of YOLO [15] has no anchor boxes, but the mproved verson YOLOv2 [16] ncorporates the dea of anchor boxes to mprove the accuracy, where the anchor shapes are obtaned by -means clusterng on the szes of the ground truth boundng boxes. When applyng the general object detectors on specfc domans, the anchor shape has to be manually modfed to mprove the accuracy. For text detecton n [8], the aspect ratos also nclude 5 : 1 and 1 : 5, snce the text could exhbt wder or hgher than the general objects. For the face detecton n [14, 24], the aspect rato s only 1 : 1 snce the face s roughly n a square shape. Once the anchor shapes are determned, the sze wll be fxed durng tranng. Ths mght be sub-optmal snce t dsregards the augmented data dstrbuton n tranng, the characterstcs of the neural networ structure and the tas. Improper desgn of the anchor sze could lead to nferor performance for specfc domans. To address the ssue, we propose a novel anchor optmzaton approach that can automatcally learn the anchor shapes durng tranng. Ths could leave the choce of anchor shapes completely n networ learnng such that the learned shapes can adapt better to the dataset, networ and tas wthout much human nterference. The learnng approach can be easly mplemented n the stochastc gradent descent way and could be plugged nto any anchor box based detecton framewor. To verfy the deas, we conduct extensve experments on several benchmar datasets ncludng Pascal VOC 07+12, MS COCO and Branwash. The results strongly demonstrate that the optmzed anchor 1

2 boxes could sgnfcantly mprove the accuracy ( 1% map absolute gan) over the baselne method. Meanwhle, the robustness s also verfed towards dfferent anchor box ntalzaton, whch greatly smplfes the problem of how to desgn the anchor sze. The man contrbutons of ths paper are summarzed as follows: We present a novel approach to optmze the anchor shapes durng tranng, whch, to the best of our nowledge, s the frst tme to treat anchor shapes as tranable varables wthout modfyng the nference networ. We demonstrate through extensve experments that the proposed anchor optmzaton method not only learns the approprate anchor shapes but also boost the detecton accuracy of exstng detectors sgnfcantly. We also verfy that the proposed method s robust towards ntalzaton, so the burden of handcraftng good anchor shapes for specfc dataset s greatly smplfed. The rest of the paper s organzed as follows. In Sec. 2, we summarze the related wors and present the relatonshp wth our approach. In Sec. 3, we present the detals of the optmzed anchor boxes for object detecton, whch s followed by the experment study n Sec. 4. Sec. 5 concludes the paper and dscusses the extensons to our wor. 2. Related Wor The modern object detectors normally contan two heads: one s the classfcaton whle the other s the localzaton. The classfcaton part s to predct the class confdence, whle the localzaton part s to predct the boundng box coordnates. Based on how the locaton s predcted, we roughly categorze the related wor nto two branches: relatve offset predcton based on some pre-defned anchor boxes [20, 13], and absolute offset predcton [15, 21, 7] Relatve Offset Predcton The networ predcts the offset relatve to the pre-defned anchor boxes, whch s also named as default boxes [13], prors [20]. These boxes serve as the ntal guess of the boundng box poston. The anchor shapes are fxed durng tranng and the neural networ learns to regress the relatve offsets. Assume ( (x), (y), (w), (h) ) are the neural net outputs, one typcal approach [18, 13] s to express the predcted boundng box as (a (x) + (x) a (w), a (y) + (y) a (h) ), a (w) exp( (w) ), a (h) exp( (h) )) where a (w) and a (h) are the pre-defned anchor wdth and heght, a (x) and a (y) are the anchor box center, the frst two numbers represent the box center and the last two represent the boundng box wdth and heght. Thus, one of the crtcal problems s how to desgn the anchor shape. In Faster R-CNN [18], the anchor shapes are chosen wth 3 scales (128 2, 256 2, ) and 3 aspect ratos (1 : 1, 1 : 2, 2 : 1), yeldng 9 dfferent anchors at each output sldng wndow poston. In Sngle Shot MultBox detector (SSD) [13], the anchor boxes also have several scales on dfferent feature map levels and aspect ratos nclude 1 : 3, 3 : 1 as well as 1 : 1, 1 : 2, 2 : 1. In YOLO [15], the networ predcts the absolute offset and has no anchor boxes, but the mproved verson of YOLOv2 [16] ncorporates the dea of anchor boxes to mprove the accuracy. The anchor shapes are not handcrafted, but are the -Means centrods wth IoU as the smlarty crteron. The utlzaton of anchors has greatly mproved deep learnng based object detecton performance n recent years. When the general object detecton framewor s appled to specfc problems, the anchor szes have to be revsted and modfed accordngly. For example of the text detecton n [8], the aspect ratos also nclude 5 : 1 and 1 : 5 as well as 1 : 1, 1 : 2, 2 : 1, 1 : 3, 3 : 1, snce the text could exhbt wder or hgher than the general objects. For the face detecton n [14, 24], the aspect rato only nclude 1 : 1 snce the face s roughly n a square shape. For pedestran detecton n [23], a rato of 0.41 based on [2] s adopted for the anchor box. As suggested n [23], napproprate anchor boxes could be nosy and degrade the accuracy. To ease the effort of anchor shape desgn, the most relevant wor mght be MetaAnchor [22]. Leveragng neural networ weght predcton, the anchors are modeled as functons mplemented by an extra neural networ and computed from customzed pror boxes. The mechansm s shown to be robust to anchor settngs and boundng box dstrbutons, compared to predefned fxed anchor scheme. However, the method nvolves an extra networ to predct the weghts of another neural networ, resultng extra tranng effort and nference tme cost, and also needs to choose a set of customzed pror boxes by hand. Comparatvely, our method can be easly embedded nto any detecton framewor wthout extra networ, and has neglgble mpact on the tranng tme/space cost and no mpact on the nference tme Absolute Offset Predcton Another research effort s to drectly predct the absolute locaton values rather than ts poston and sze relatve to pre-defned anchor boxes. The YOLO [15] belongs to ths spectrum but was mproved by YOLOv2 [16] wth anchorbased approach. For DeNet [21], the networ outputs the confdence of each neuron belongng to one of the boundng box corners, and then collects the canddate boxes by Drected Sparse Samplng. More recently, CornerNet [7] proposed detectng objects by the top-left and bottom-rght eypont pars, and ntroduces the corner poolng operaton to better localze corners. Whle these two anchor-

3 free methods form a promsng future research drecton, yet anchor-based methods stll acheves the best accuracy n the publc benchmars. 3. Proposed Approach We frst present an overvew on exstng anchor-based object detecton framewors, and then descrbe the proposed optmzaton technques n detals Object Detecton Overvew In state-of-the-art object detecton framewors, the tranng procedure s normally formulated as an emprcal mnmzaton problem over a combnaton of boundng box localzaton loss and the classfcaton loss Localzaton Loss For one feature map wth A dfferent anchor shapes from the networ, each spatal locaton could correspond to A anchor boxes centered at the cell. Thus the total number of anchor boxes are N A H f W f, where H f and W f are the feature map heght and wdth, respectvely. Stacng all the anchor boxes, we can denote by a = (a (x), a (y), a (w), a (h) ) the -th ( {1,, N}) anchor box, where a (x) and a (y) represents the center of the box and a (w) and a (h) represent the wdth and heght, respectvely. For multple feature maps as n [11, 13], we can also use smlar notatons to represent all the anchor boxes staced together. Note snce we have A dfferent anchor shapes, the value of a (w) and a (h) can have A dfferent values nstead of N dfferent values. The anchor center of a (x) and a (y) are normally lnearly related the spatal locaton n the feature map. The shape a (w) and a (h) and reman constant durng tranng n exstng wor. are pre-defned Let = ( (x), (y), (w), (h) ) be the networ output for the -th anchor box. Then, the localzaton loss s to algn the networ predcton to the ground-truth boundng box coordnates wth respect to the anchor box. Specfcally, the loss for the -th anchor box could be wrtten as where g j = (g (x) j L loc = δ,j L( ; a, g j ), (1), g (y) j, g (w) j, g (h) j ) are the j-th ground-truth box and δ,j measures how much the -th anchor should be responsble to the j-th ground-truth. The value of δ,j s usually restrcted to dscrete value n {0, 1}, n whch 1 ndcates that -th anchor box s responsble for the j-th ground-truth box. For example n [18, 11, 13], δ,j s 1 f the IoU rato between the anchor box and the ground-truth box s larger than a threshold e.g. 0.5 or the anchor box s the one wth the largest overlap wth the ground-truth. In YOLOv2 [16], δ,j s 1 f the anchor box and the ground-truth are located n the same spatal locaton and the anchor box s the one wth the largest IoU wth the ground-truth box. The form of the localzaton loss could be the L 2 dstance [16], or the smoothed L 1 loss (also nown as Huber loss) [18, 13]. Tang the L 2 loss as the example, the loss of L( ; a, g j ) can be wrtten as the sum of L (x,y),j wth L (w,h),j L (x,y),j L (w,h),j â (w) ĝ (w) j =( (x) =( (w) + a (x) + â (w) and g (x) j ) 2 + ( (y) + a (y) g (y) j ) 2 (2) ĝ (w) j ) 2 + ( (h) + â (h) ĝ (h) j ) 2 (3) log(a (w) ), â (h) log(a (h) ) (4) log(g (w) j ), ĝ (h) j log(g (h) j ) (5) The wdth and heght are wth the log encodng scheme because the value should always be postve. Note that they appear explctly n the wh-loss term Eqn. 3. Ths enables drect gradent computaton on a (w) j, a (h) j, whch s the ey of our anchor optmzaton method and wll be detaled n Sec Classfcaton Loss For each anchor box, the networ also outputs the confdence score to dentfy whch class t belongs to. In tranng, normally cross entropy loss s employed, e.g. n [18, 13, 15, 16]. One mproved verson s the focal loss [11], whch focuses on the mbalance ssue. To handle the bacground class, one can use an extra bacground class n the cross entropy loss, e.g. n [13, 18]. Another approach s to learn a class-agnostc objectness score to dentfy f there s an object, e.g. n YOLOv2[16] and the RPN of Faster R-CNN[18] Anchor Box Optmzaton By combnng the localzaton loss and the classfcaton loss, we can wrte the optmzaton problem as θ,{(s (w),s(h) )}A =1 mn θ L(θ) (6) where θ s the neural networ parameters. In exstng methods, the anchor shapes are treated as constants. For all the N anchor boxes a, we extract all the dstnct anchor shapes and denote them by (s (w), s(h) )A =1. We propose to treat them as learnable varables n the optmzaton problem Eqn. 7. ( ) mn L θ, {(s (w), s(h) )}A =1 (7)

4 Gradent Classfcaton Loss CNN s (w) s (h) Anchor Shapes (x) (y) (w) (h) (x) (y) (w) (h) Δ 1 Δ1 Δ1 Δ1 Δ Δ Δ Δ (x) (y) (w) (h) (x) (y) (w) (h) a 1 a1 a1 a1 a a a a Gradent + Gradent Localzaton Loss δ,j L Δ ; a, g j j Onlne Clusterng Warm-Up Soft Assgnment Warm-Up Batch Normalzaton w/o Shftng Fgure 1. An llustraton of the anchor optmzaton process. The localzaton loss s to mnmze the error between the ground-truth boundng box and the predcted offset relatve to the anchor box. The error s bac-propagated to the anchor shapes as well as the CNN parameters to automatcally learn the anchor sze. The anchor shape s warmed up by the onlne clusterng and the soft assgnment wth the batch normalzaton wthout shftng. Obvously, Eqn. 7 s guaranteed to reach a lower optmal loss value than Eqn. 6 snce the set of learnable varable set s enlarged (so s the feasble soluton set). The anchor shape values can be adjusted n the goal of lowerng the overall loss value. Moreover, wth the learned optmal anchor shapes, the magntudes of the offsets (resduals) become smaller, whch mght mae the regresson problem easer. The ey dea s summarzed n Fg. 1. Followng common practce, we use the bac-propagaton to solve the optmzaton problem. Instead of learnng s w and sh, we learn ŝ w log(sw ) and ŝh log(sh ) because of equvalence and smplcty. For one tranng mage, the dervatve of the loss functon wth respect to ŝ (w) can be computed as L ( ) δ ŝ (w),j L (w,h),j ŝ (w),j (8) ( δ,j + â w ĝ w ) j δ(â w = ŝ w ), (9),j where δ(â w = ŝ w ) s 1 f âw corresponds to â w j and 0, otherwse. Smlarly, we can have the dervatve wth respect to the anchor heght â (h). In one tranng teraton of the mn-batch stochastc gradent descent algorthm, we frstly assgn the ground-truth boxes to the anchors,.e. computes δ,j. Then, wth δ,j fxed, bac-propagate the error sgnal to all remanng parameters ncludng the anchor shapes. To calculate the varables δ,j, we normally use the IoU as the metrc [18, 13, 16]. If we use L 2 dstance n the log space of wdth and heght as dstance metrc to compute δ,j, the method algns more closely wth the loss. Emprcally, we fnd that usng L 2 dstance or IoU results n smlar performance and anchor shapes. To further facltate automatc learnng of anchor shapes, we ntroduce the followng three tranng technques Onlne Clusterng Warm-Up Motvated by the -means approach n [16], we augment the loss functon wth an extra onlne clusterng term durng the early stage of tranng. Ths term mnmzes the squared L 2 dstance between the anchor shapes and the ground-truth box shapes and can be wrtten as L aug = L + λ 1 2N T,j (â (w) δ,j T,j, (10),j where N s,j δ,j for normalzaton. ĝ (w) j ) 2 + (â (h) ĝ (h) j ) 2, (11) The coeffcent λ s lnearly annealed from 1 to 0 durng the early stage of tranng (frst 20% teratons n experments) to c off the learnng of anchors. The underlyng dea s that the -means approach could serve as a good startng pont. Ths maes the networ more robust to the ntalzaton and fast to converge. In the early tranng stage, the clusterng term could qucly tune the anchors to (near) -means centrods. Then, the orgnal loss of L n Eqn. 7 begns to show more nfluence. Hence, the anchor shapes adapt more closely to the data dstrbuton and the networ predctons, followng gradents comng from the orgnal loss term. The dervatves of the augmented loss n Eqn. 10 wth

5 respect to ŝ (w) and ŝ (h) are L aug ŝ (w) L aug ŝ (h) = L + λ ŝ (w) N = L + λ ŝ (h) N,j,j δ,j (â (w) δ,j (â (h) Soft Assgnment Warm-Up ĝ (w) j )δ(â (w) = ŝ (w) ) ĝ (h) j )δ(â (h) = ŝ (h) ) In some extreme stuaton, the ground-truth boundng box could be very small or very large, and only one anchor box s actvated. All other anchor boxes are never used, even f we have the onlne clusterng term. To address the ssue, we propose to adopt a soft assgnment approach at the early tranng stage. That s δ,j = Softmax( dst(a, g j )/T ), (12) where Softmax s the softmax functon over all anchor boxes at the same spatal cell. The temperature T s annealed from 2 to 0 n the frst few tranng steps (1500 teratons n the experments). Wth non-zero assgnment values, all anchor shapes could jon nto the learnng procedure. After the warm-up, t falls bac to the orgnal assgnment scheme. In the normal tranng tas, we fnd ths tem has almost no effect on the accuracy, but n specfc tas doman, t sgnfcantly solves the problem and mproves the accuracy Batch Normalzaton wthout Shftng Wth the onlne clusterng term n Eqn. 11, the networ output tends to have a zero mean potentally followng Gaussan dstrbuton. To further reduce the learnng dffcultes, we apply the batch normalzaton [5] on the output of w and h wthout the shftng parameters. That s, the networ output s frst normalzed to zero mean and unt varance, followed by scalng operaton wthout the shft operaton. Ths could enforce the zero mean dstrbuton and mae the networ converge fast. 4. Experments We frst present the mplementaton detals and then the extensve experment results on wdely used Pascal VOC [3] and MS-COCO Challenge 2017 object detecton datasets [12], along wth a head detecton dataset named Branwash [19], to demonstrate the effectveness of the proposed anchor optmzaton method Implementaton Detals Snce the proposed approach for optmzng anchors s qute general, t can be appled to most anchor-based object detecton framewors. We choose the YOLOv2 [16] framewor as the testbed to demonstrate the effectveness. Extensons to other detectors should be straghtforward, such as the RPN of Faster R-CNN [18], SSD [13], Feature Pyramd Networ (FPN) [9], and RetnaNet [11]. YOLOv2 s one of the typcal one-stage detectors, whch maps the nput mage to a feature map by convolutonal neural networ and nfers the boundng box relatve offsets and the classfcaton results based on the feature map. The networ conssts of a DarNet-19 bacbone CNN pretraned on ImageNet classfcaton dataset, and several convolutonal detecton heads. Wth A = 5 anchor shapes, the last convolutonal layer outputs a feature map of 5 ( C) channels, correspondng to 4 coordnate regresson outputs ( ), 1 class-agnostc objectness score, and C category scores for each anchor box. We also employ the same data augmentaton technques as n YOLOv2, ncludng random jtterng, scalng, and random hue, exposure, saturaton change of the mage. The same loss weghts are used to balance the localzaton loss, the objectness loss and the classfcaton loss. Durng testng, an mage s reszed to a specfed sze (e.g. 416-by-416 pxels), and then fed nto the detecton networ. For each anchor box a and the correspondng output, the output boundng box s (a (x) (y), a (w) exp{ (w) }, a (h) exp{ (h) + (x), a (y) + }) wth the score beng the multplcaton of the objectness score and the condtonal classfcaton score. The fnal predcton results are the top- (typcally = 300) canddate boxes sorted by the box scores, after the class specfc Non-Maxmum Suppresson (NMS) wth IoU threshold as We mplement the approach on Caffe [6] Experment Results PASCAL VOC The PASCAL VOC dataset [3] contans box annotatons over 20 object categores. We adopt the commonly used tran/test splt, where the VOC 2007 tranval (5 mages) and VOC 2012 tranval (11 mages) are used as tranng set, and VOC 2007 test (4952 mages) s used as testng set. The model tranng s done n 30,000 teratons of SGD (Momentum = 0.9) wth mn-batch sze 64 evenly dvded onto 4 GPUs. The learnng rate s set to step-wse schedule: (0 100,1e-4), (100 15,000,1e-3), (15,000 27,000,1e-4), (27,000 30,000, 1e-5). The tranng mage sze s set to 416 or 544 to match the test sze. The anchor shapes are ntalzed by three methods to study the robustness. 1. unform: The anchor shapes are chosen unformly,.e. [(3, 3), (3, 9), (9, 9), (9, 3), (6, 6)] strde wth the strde beng 32 here.

6 Table 1. Detecton results on Pascal VOC 2007 test set, traned on VOC tranval sets. Sze represents the shorter edge of test mage sze. map.5 stands for mean average precson at IoU 0.5. AP for each class s also reported. Method Sze map.5 aero be brd boat bottle bus car cat char cow table dog horse mbe person plant sheep sofa tran tv Faster rcnn vgg[18] Faster rcnn res[4] SSD512 [13] YOLOv2 [16] YOLOv2 [16] Baselne (dentcal) Baselne (unform) Baselne (-means) Baselne (-means) Opt (dentcal) Opt (unform) Opt (-means) Opt (-means) dentcal: All 5 anchors boxes are dentcal and ntalzed as (5, 5) strde. 3. -means: The values are borrowed from the open source code of YOLOv2 1 to perform the -means clusterng on the ground-truth box shapes wth the IoU as metrc. 1 Fgure 2. Pascal VOC anchors and box dstrbuton n log scale. The red star marers show the learned anchor shapes. Underlyng the marers s Kernel Densty of the boundng box wth and heght wth the mge reszed to Darer color means hgher densty. Around the fgure are the margnal dstrbutons of log(w) and log(h). The results are shown n Table 1. We also lst Faster R- CNN and SSD results n the table for completeness. Note that we are not targetng the best accuracy but manly the effectveness of the proposed approach. In the Baselne (*) rows of the table, the anchor shapes are fxed as n conventonal detecton model tranng, whereas n the Opt (*) rows, the anchor shapes are optmzed wth our proposed method n Sec. 3. From the results, our anchor optmzaton method consstently produces better results compared to the baselnes. Our re-mplementatons of YOLOv2 attan smlar or better performances compared to what the orgnal paper reports (comparng Baselne (-means) wth YOLOv2). The proposed anchor learnng method further boosts the performance by more than 1.2% n terms of absolute map value. For example wth -means ntalzaton and 544 as the mage sze, the baselne acheves 79.45% map, whle our method boosts the accuracy to 80.69%, leadng to 1.2 pont mprovement. Furthermore, dfferent anchor shape ntalzaton acheves smlar accuracy. Wthn our proposed approaches wth dfferent ntalzatons, the accuracy dfference between the best and the worse s only 0.06 pont for the sze of 416, suggestng that our method s very robust to dfferent ntal anchor confguratons. Hence, the manual choce of approprate ntal anchor shapes becomes less crtcal wth our method. Note for the settng of dentcal, though the anchor szes are the same at the begnnng, the values can be optmzed to dfferent values snce dfferent anchors are responsble for dfferent ground-truth boxes durng tranng. Fgure 2 llustrates the unform anchors, the -means anchors, the learned anchors (wth -means ntalzaton), and the ground-truth box shape dstrbuton n the log w-log h plane. We observe that both the learned anchors and the -means anchors algn closely wth the underlyng ground truth box dstrbuton, whch ntutvely explans why they

7 produce better performance than the unform anchors. The learned anchors spread broader and are slghtly smaller than the -means anchors. The reason mght be that the small boundng box s relatvely hard to regress and the networ pushes the anchors to focus more on small objects to lower the loss. Ths ndcates that the anchor optmzaton process s more than merely clusterng. It s also able to adapts the anchor shapes to the data augmentaton and the networ regresson capablty to mprove the accuracy MS COCO Table 2. Detecton results on MS COCO val. Average Precsons (AP) at dfferent IoU thresholds and dfferent box scales (Small, Medum, Large at IoU 0.5) are reported. Method AP.5:.95 AP.5 AP.75 AP.5S AP.5M AP.5L Baselne (unform) Baselne (-means) Opt (dentcal) Opt (unform) Opt (-means) We adopt the frequently used COCO [12] 2017 Detecton Challenge tran/val splts, where the tranng set has 115K mages, the val set has 5K mages, and the test-dev set contans about 20 mages whose box annotatons are not publcly avalable. The dataset contans 80 object categores. We use smlar tranng confguratons as the VOC experments. Mn-batch sze s 64 and evenly splt n 4 GPUs. Momentum of SGD s set to 0.9. Snce the COCO dataset has substantally more mages than VOC, we ncrease the number of teratons to 100,000, and set the learnng rate schedule to (0 1,000,1e-4), (1,000 80,000,1e- 3), (80,000 90,000,1e-4), (90, ,000, 1e-5). The tranng and testng mage szes are both set to 544 n all experments. Snce the boundng box annotatons of the test-dev s not exposed, we upload our detecton results to the offcal COCO evaluaton server 2 to retreve the scores. The results on the val set are shown n Table 2, and the results on the test-dev set are n Table 3. In the table, AP.5:.95 denotes the mean of AP evaluated at IoU threshold evenly dstrbuted between 0.5 and 0.95; AR denotes the average recall rate. Compared to the orgnal YOLOv2 results, our remplementaton even acheves a hgher accuracy wth the AP.5:.95 ncreased from 21.6% to 24.0% and AP.5 ncreased from 44.0% to 44.9%. When equpped wth the proposed anchor optmzaton method, the accuracy s further sgnfcantly mproved by 1%, wth AP.5:.95 to 25.0% and 2 Fgure 3. MS COCO anchors and box dstrbuton n log scale. Underlyng the marers s Kernel Densty of the ground-truth boundng box wdth and heght wth the mage reszed to 544x544. Around the fgure are the margnal dstrbutons of log(w) and log(h). AP.5 to 45.9%. Smlar mprovements can also be observed from the results on val splt. Ths strongly demonstrates the superor of the anchor optmzaton to acheve hgher accuracy. Meanwhle, the baselne approach wthout anchor optmzaton s qute senstve to the anchor shapes. On val, -means ntalzaton acheves 23.45%, whle the unform ntalzaton acheves 21.90% wth 1.55 pont dfference. Comparatvely, our optmzaton approach s more robust and the dfference between the hghest (24.55 on val ) and the lowest (24.43 on val) s only On test-dev, dfferent ntalzaton methods acheve the same map.5:.95 (25.0), whch further verfes the robustness towards dfferent ntalzaton methods. The learned anchors wth dfferent ntalzatons are shown n Table 4, and we can easly observe that the anchor shapes are qute smlar though the ntal values are dfferent. Fgure 3 shows the learned anchor shapes aganst the unform and the -means anchors. The learned anchors ncely cover the ground-truth boundng box dstrbuton. They also tend to be slghtly smaller than the orgnal -means values, whch could help the small object detecton snce the large object s relatvely easy to detect. Ths can also be verfed from Table 3. Tang the nstance of -means ntalzaton, the gan from small (from 4.4% to 5.7%) and medum object (from 24.6% to 26.6%) s hgh whle t even sacrfces the accuracy for large objects a lttle bt (from 40.9% to 40.8%).

8 Table 3. Detecton results from the evaluaton server on MS COCO test-dev. AP means average precson, AR means average recall. AP.5:.95 s the mean of AP at IoU 0.5:0.05:0.95. Subscrpt S,M & L correspond to small, medan & large boundng boxes respectvely. Method AP.5:.95 AP.5 AP.75 AP S AP M AP L AR 1 AR 10 AR 100 AR S AR M AR L Faster RCNN vgg[18] Faster RCNN [1] SSD512 [13] YOLOv2 [16] Baselne (unform) Baselne (-means) Opt (dentcal) Opt (unform) Opt (-means) Table 4. Learned anchors from dfferent ntalzatons on COCO wth mage sze as 544. Int s (w) 1, s (h) 1 s (w) 2, s (h) 2 s (w) 3, s (h) 3 s (w) 4, s (h) 4 s (w) 5, s (h) 5 dentcal 5.8, , , , , 237 unform 5.8, , , , , 237 -means 5.7, , , , , Branwash Branwash s a head detecton dataset ntroduced n [19], whch has about 10 mages for tranng, about 500 mages for valdaton and 484 mages for testng. The mages are of ndoor scenes where people come and go captured wth a survellance camera. We tran the detecton model for 10,000 steps, wth learnng rate schedule (0 100,1e-4), (100 5,000,1e-3), (5,000 9,000,1e-4), (9,000 10,000, 1e-5). No random scalng augmentaton s used snce the camera s stll, whle other nds of data augmentaton reman unchanged. The mage crop sze durng tranng s set to 320, and the test mage sze s chosen as 640. We stll choose to employ 5 anchor shapes. No classfcaton loss s appled snce there s only one category (head). We report AP.5 as the performance crteron n Table 5. The baselne result wth the anchor shapes from COCO s also presented. The -means anchors are computed n smlar way as n YOLOv2. Snce the head boundng boxes are much smaller than those of the VOC or COCO datasets, we fnd that only one anchor shape wll be actvated throughout the tranng and the remanng anchor shapes never get used wth the anchor shape from COCO settngs. In ths case, the neural networ wll also need to predct large devatons for w and h to ft all the ground-truth boxes, whch s suboptmal. Ths means t s sub optmal to use the anchor shapes from other domans. Comparably, the proposed anchor learnng method could adjust the anchor shape qucly to cover the ground-truth boundng box well. From the results, we can observe Opt (*) consstently outperform the baselnes by a large margn, demonstratng the effectveness of the proposed method. Even wth the -means as the ntalzed anchor shapes, our approach can also mprove the accuracy by 1.2 pont (from 78.98% to 80.18%). Table 5. Detecton results on Branwash dataset. Test mage sze s 640. AP.5 s the average precson wth IoU threshold Concluson Method Sze AP.5 Baselne (coco) Baselne (unform) Baselne (-means) Opt (dentcal) Opt (unform) Opt (-means) In ths paper, we have ntroduced an anchor optmzaton method whch can be employed n most exstng anchorbased object detecton framewors to automatcally learn the anchor shapes durng tranng. The learned anchors are better suted for specfc data and networ structure and can produce better accuracy. We demonstrated the effectveness of the proposed method based on the popular one-stage object detecton framewor YOLOv2. Extensve experments on Pascal VOC, MS COCO and Branwash benchmar datasets show superor detecton accuracy of our proposed method over the baselne. We also show that the anchor optmzaton method s robust to ntalzaton (dentcal, unform, -means), and hence the careful handcraftng of anchor shapes s greatly allevated for good performance. Moreover, the proposed method s qute general. The same method can also be appled n other one-stage methods such as SSD[13], RetnaNet[11], etc., whch s based on the anchor box, and n two-stage methods to mprove the regon proposals. The method s ndependent to mprove-

9 ments such as Feature Pyramds Networ (FPN) [10] and thus can potentally be combned wth them to further boost performance. Our wor solves the problem of optmzng anchor shapes, but not of the number of anchors, whch would be an nterestng topc to study. Fnally, theoretcal wors on why and how the anchor mechansm wors better than plan regresson would also be very valuable to the feld. References [1] Coco: Common objects n context. org/dataset/#detectons-leaderboard. Accessed: [2] P. Dollár, C. Woje, B. Schele, and P. Perona. Pedestran detecton: An evaluaton of the state of the art. IEEE Transactons on Pattern Analyss and Machne Intellgence, 34: , [3] M. Everngham, L. V. Gool, C. K. I. Wllams, J. M. Wnn, and A. Zsserman. The pascal vsual object classes (voc) challenge. Internatonal Journal of Computer Vson, 88: , [4] K. He, X. Zhang, S. Ren, and J. Sun. Deep resdual learnng for mage recognton IEEE Conference on Computer Vson and Pattern Recognton (CVPR), pages , [5] S. Ioffe and C. Szegedy. Batch normalzaton: Acceleratng deep networ tranng by reducng nternal covarate shft. In ICML, [6] Y. Ja, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Grshc, S. Guadarrama, and T. Darrell. Caffe: Convolutonal archtecture for fast feature embeddng. arxv preprnt arxv: , [7] H. Law and J. Deng. Cornernet: Detectng objects as pared eyponts. CoRR, abs/ , [8] M. Lao, B. Sh, and X. Ba. Textboxes++: A sngle-shot orented scene text detector. IEEE Trans. Image Processng, 27(8): , , 2 [9] T. Ln, P. Dollár, R. B. Grshc, K. He, B. Harharan, and S. J. Belonge. Feature pyramd networs for object detecton. In 2017 IEEE Conference on Computer Vson and Pattern Recognton, CVPR 2017, Honolulu, HI, USA, July 21-26, 2017 [10], pages [10] T.-Y. Ln, P. Dollár, R. B. Grshc, K. He, B. Harharan, and S. J. Belonge. Feature pyramd networs for object detecton IEEE Conference on Computer Vson and Pattern Recognton (CVPR), pages , [11] T.-Y. Ln, P. Goyal, R. B. Grshc, K. He, and P. Dollár. Focal loss for dense object detecton. IEEE transactons on pattern analyss and machne ntellgence, , 3, 5, 8 [12] T.-Y. Ln, M. Mare, S. J. Belonge, L. D. Bourdev, R. B. Grshc, J. Hays, P. Perona, D. Ramanan, P. Dollár, and C. L. Ztnc. Mcrosoft coco: Common objects n context. In ECCV, , 7 [13] W. Lu, D. Anguelov, D. Erhan, C. Szegedy, S. E. Reed, C.- Y. Fu, and A. C. Berg. Ssd: Sngle shot multbox detector. In ECCV, , 2, 3, 4, 5, 6, 8 [14] M. Najb, P. Samangoue, R. Chellappa, and L. S. Davs. SSH: sngle stage headless face detector. In IEEE Internatonal Conference on Computer Vson, ICCV 2017, Vence, Italy, October 22-29, 2017, pages , , 2 [15] J. Redmon, S. K. Dvvala, R. B. Grshc, and A. Farhad. You only loo once: Unfed, real-tme object detecton IEEE Conference on Computer Vson and Pattern Recognton (CVPR), pages , , 2, 3 [16] J. Redmon and A. Farhad. Yolo9000: Better, faster, stronger IEEE Conference on Computer Vson and Pattern Recognton (CVPR), pages , , 2, 3, 4, 5, 6, 8 [17] J. Redmon and A. Farhad. Yolov3: An ncremental mprovement. CoRR, abs/ , [18] S. Ren, K. He, R. B. Grshc, and J. Sun. Faster r-cnn: Towards real-tme object detecton wth regon proposal networs. IEEE Transactons on Pattern Analyss and Machne Intellgence, 39: , , 2, 3, 4, 5, 6, 8 [19] M. A. Russell Stewart and A. Y. Ng. End-to-end people detecton n crowded scenes IEEE Conference on Computer Vson and Pattern Recognton (CVPR), pages , , 8 [20] C. Szegedy, S. E. Reed, D. Erhan, and D. Anguelov. Scalable, hgh-qualty object detecton. CoRR, abs/ , [21] L. Tychsen-Smth and L. Petersson. Denet: Scalable realtme object detecton wth drected sparse samplng IEEE Internatonal Conference on Computer Vson (ICCV), pages , [22] T. Yang, X. Zhang, W. Zhang, and J. Sun. Metaanchor: Learnng to detect objects wth customzed anchors. NIPS, abs/ , [23] L. Zhang, L. Ln, X. Lang, and K. He. Is faster r-cnn dong well for pedestran detecton? In European Conference on Computer Vson, pages Sprnger, [24] S. Zhang, X. Zhu, Z. Le, H. Sh, X. Wang, and S. Z. L. Sˆ3fd: Sngle shot scale-nvarant face detector. In IEEE Internatonal Conference on Computer Vson, ICCV 2017, Vence, Italy, October 22-29, 2017, pages , , 2

A Quantization-Friendly Separable Convolution for MobileNets

A Quantization-Friendly Separable Convolution for MobileNets arxv:1803.08607v1 [cs.cv] 22 Mar 2018 A Quantzaton-Frendly Separable for MobleNets Abstract Tao Sheng tsheng@qt.qualcomm.com Xaopeng Zhang parker.zhang@gmal.com As deep learnng (DL) s beng rapdly pushed

More information

LOW-COMPLEXITY VIDEO ENCODER FOR SMART EYES BASED ON UNDERDETERMINED BLIND SIGNAL SEPARATION

LOW-COMPLEXITY VIDEO ENCODER FOR SMART EYES BASED ON UNDERDETERMINED BLIND SIGNAL SEPARATION LOW-COMPLEXITY VIDEO ENCODER FOR SMART EYES BASED ON UNDERDETERMINED BLIND SIGNAL SEPARATION Jng Lu, Fe Qao *, Zhjan Ou and Huazhong Yang Department of Electronc Engneerng, Tsnghua Unversty ABSTRACT Ths

More information

Following a musical performance from a partially specified score.

Following a musical performance from a partially specified score. Followng a muscal performance from a partally specfed score. Bryan Pardo and Wllam P. Brmngham Artfcal Intellgence Laboratory Electrcal Engneerng and Computer Scence Dept. and School of Musc The Unversty

More information

The UCD community has made this article openly available. Please share how this access benefits you. Your story matters!

The UCD community has made this article openly available. Please share how this access benefits you. Your story matters! Provded by the author(s) and Unversty College Dubln Lbrary n accordance wth publsher polces., Please cte the publshed verson when avalable. tle Dynamc Complexty Scalng for Real-me H.264/AVC Vdeo Encodng

More information

Instructions for Contributors to the International Journal of Microwave and Wireless Technologies

Instructions for Contributors to the International Journal of Microwave and Wireless Technologies Instructons for Contrbutors to the Internatonal Journal of Mcrowave and Wreless Technologes Frst A. Author 1, Second Author 1,2, Thrd Author 2 1 Cambrdge Unversty Press, Ednburgh Buldng, Shaftesbury Road,

More information

A Comparative Analysis of Disk Scheduling Policies

A Comparative Analysis of Disk Scheduling Policies A Comparatve Analyss of Dsk Schedulng Polces Toby J. Teorey and Tad B. Pnkerton Unversty of Wsconsn* Fve well-known schedulng polces for movable head dsks are compared usng the performance crtera of expected

More information

Error Concealment Aware Rate Shaping for Wireless Video Transport 1

Error Concealment Aware Rate Shaping for Wireless Video Transport 1 Error Concealment Aware Rate Shapng for Wreless Vdeo Transport 1 Trsta Pe-chun Chen and Tsuhan Chen 2 Abstract Streamng of vdeo, whch s both source- and channel- coded, over wreless networks faces the

More information

current activity shows on the top right corner in green. The steps appear in yellow

current activity shows on the top right corner in green. The steps appear in yellow Browzwear Tutorals Tutoral ntroducton Ths tutoral leads you through the best practces of color ways operatons usng an llustrated step by step approach. Each slde shows the actual applcaton at the stage

More information

Statistics AGAIN? Descriptives

Statistics AGAIN? Descriptives Cal State Northrdge Ψ427 Andrew Answorth PhD Statstcs AGAIN? What do we want to do wth statstcs? Organze and Descrbe patterns n data Takng ncomprehensble data and convertng t to: Tables that summarze the

More information

Optimized PMU placement by combining topological approach and system dynamics aspects

Optimized PMU placement by combining topological approach and system dynamics aspects Optmzed PU placement by combnng topologcal approach and system dynamcs aspects Jonas Prommetta, Jakob Schndler, Johann Jaeger Insttute of Electrcal Energy Systems Fredrch-Alexander-Unversty Erlangen-Nuremberg

More information

Accepted Manuscript. An improved artificial bee colony algorithm for flexible job-shop scheduling problem with fuzzy processing time

Accepted Manuscript. An improved artificial bee colony algorithm for flexible job-shop scheduling problem with fuzzy processing time Accepted Manuscrpt An mproved artfcal bee colony algorthm for flexble ob-shop schedulng problem wth fuzzy processng tme Ka Zhou Gao, Ponnuthura Nagaratnam Suganthan, Quan Ke Pan, Tay Jn Chua, Chn Soon

More information

System of Automatic Chinese Webpage Summarization Based on The Random Walk Algorithm of Dynamic Programming

System of Automatic Chinese Webpage Summarization Based on The Random Walk Algorithm of Dynamic Programming Send Orders for Reprnts to reprnts@benthamscence.ae The Open Cybernetcs & Systemcs Journal, 205, 9, 35-322 35 Open Access System of Automatc Chnese Webpage Summarzaton Based on The Random Walk Algorthm

More information

arxiv: v1 [cs.cl] 12 Sep 2018

arxiv: v1 [cs.cl] 12 Sep 2018 Powered by TCPDF (www.tcpdf.org) Neural Melody Composton from Lyrcs Hangbo Bao, Shaohan Huang 2, Furu We 2, Le Cu 2, Yu Wu 3, Chuanq Tan 3, Songhao Pao, Mng Zhou 2 School of Computer Scence, Harbn Insttute

More information

The Traffic Image Is Dehazed Based on the Multi Scale Retinex Algorithm and Implementation in FPGA Cui Zhe1, a, Chao Li2, b *, Jiaqi Meng3, c

The Traffic Image Is Dehazed Based on the Multi Scale Retinex Algorithm and Implementation in FPGA Cui Zhe1, a, Chao Li2, b *, Jiaqi Meng3, c 3rd Internatonal Conference on Mechatroncs and Industral Informatcs (ICMII 2015) The Traffc Image Is Dehazed Based on the Mult Scale Retnex Algorthm and Implementaton n FPGA Cu Zhe1, a, Chao L2, b *, Jaq

More information

tj tj D... '4,... ::=~--lj c;;j _ ASPA: Automatic speech-pause analyzer* t> ,. "",. : : :::: :1'NTmAC' I

tj tj D... '4,... ::=~--lj c;;j _ ASPA: Automatic speech-pause analyzer* t> ,. ,. : : :::: :1'NTmAC' I ASPA: Automatc speech-pause analyzer* D. GERVERt and G. DNELEY Unversty of Durham, Durham, England ASPA: The Programs Snce the actual detals of nterface samplng, dsk storage routnes, etc., wll depend upon

More information

Novel Quantization Strategies for Linear Prediction with Guarantees

Novel Quantization Strategies for Linear Prediction with Guarantees Smon S. Du* Ychong Xu* Yuan L Hongyang Zhang Aart Sngh Pulkt Grover Carnege Mellon Unversty, Pttsburgh, PA 15213, USA *: Contrbute equally. SSDU@CS.CMU.EDU YICHONGX@CS.CMU.EDU LIYUANCHRISTY@GMAIL.COM HONGYANZ@CS.CMU.EDU

More information

Technical Information

Technical Information CHEMCUT Techncal Informaton CORPORATION Introducton The Chemcut CC8000 etcher has many new features desgned to reduce the cost of manufacturng and, just as mportantly, the cost of ownershp. Keepng the

More information

Cost-Aware Fronthaul Rate Allocation to Maximize Benefit of Multi-User Reception in C-RAN

Cost-Aware Fronthaul Rate Allocation to Maximize Benefit of Multi-User Reception in C-RAN Cost-Aware Fronthaul Rate Allocaton to Maxmze Beneft of Mult-User Recepton n C-RAN Dora Bovz, Chung Shue Chen, Sheng Yang To cte ths verson: Dora Bovz, Chung Shue Chen, Sheng Yang. Cost-Aware Fronthaul

More information

Hybrid Transcoding for QoS Adaptive Video-on-Demand Services

Hybrid Transcoding for QoS Adaptive Video-on-Demand Services 732 IEEE Transactons on Consumer Electroncs, Vol. 50, No. 2, MAY 2004 Hybrd Transcodng for QoS Adaptve Vdeo-on-Demand Servces Ilhoon Shn and Kern Koh Abstract Transcodng s a core technque that s used n

More information

Simon Sheu Computer Science National Tsing Hua Universtity Taiwan, ROC

Simon Sheu Computer Science National Tsing Hua Universtity Taiwan, ROC Mounr A. Tantaou School of Electrcal Engneerng and Computer Scence Unversty of Central Florda Orlando, FL 3286-407-823-393 tantaou@cs.ucf.edu Interacton wth Broadcast Vdeo Ken A. Hua School of Electrcal

More information

Why Take Notes? Use the Whiteboard Capture System

Why Take Notes? Use the Whiteboard Capture System Why Take Notes? Use the Whteboard Capture System L-we He Zhengyou Zhang and Zcheng Lu September, 2002 Techncal Report MSR-TR-2002-89 Mcrosoft Research Mcrosoft Corporaton One Mcrosoft Way Redmond, WA 98052

More information

A Scalable HDD Video Recording Solution Using A Real-time File System

A Scalable HDD Video Recording Solution Using A Real-time File System H. L et al.: A Scalable HDD Vdeo Recordng Soluton Usng A Real-tme Fle System A Scalable HDD Vdeo Recordng Soluton Usng A Real-tme Fle System Hong L, Stephen R. Cumpson Member, IEEE, Robert Jochemsen, Jan

More information

Decision Support by Interval SMART/SWING Incorporating. Imprecision into SMART and SWING Methods

Decision Support by Interval SMART/SWING Incorporating. Imprecision into SMART and SWING Methods Decson Support by Interval SMART/SWING Incorporatng Imprecson nto SMART and SWING Methods Abstract: Interval judgments are a way of handlng preferental and nformatonal mprecson n multcrtera decson analyss.

More information

QUICK START GUIDE v0.98

QUICK START GUIDE v0.98 QUICK START GUIDE v0.98 QUICK HELP Q A 1 STEP BY STEP 3 GLOSSARY 2 A B C 1 INSTALLATION 1. Make sure that the hardware nstallaton s performed by a certfed vendor 2. Install OTOTRAK app from Apple s App

More information

SONG STRUCTURE IDENTIFICATION OF JAVANESE GAMELAN MUSIC BASED ON ANALYSIS OF PERIODICITY DISTRIBUTION

SONG STRUCTURE IDENTIFICATION OF JAVANESE GAMELAN MUSIC BASED ON ANALYSIS OF PERIODICITY DISTRIBUTION SOG STRUCTURE IDETIFICATIO OF JAVAESE GAMELA MUSIC BASED O AALYSIS OF PERIODICITY DISTRIBUTIO D. P. WULADARI, Y. K. SUPRAPTO, 3 M. H. PUROMO,,3 Insttut Teknolog Sepuluh opember, Department of Electrcal

More information

TRADE-OFF ANALYSIS TOOL FOR INTERACTIVE NONLINEAR MULTIOBJECTIVE OPTIMIZATION Petri Eskelinen 1, Kaisa Miettinen 2

TRADE-OFF ANALYSIS TOOL FOR INTERACTIVE NONLINEAR MULTIOBJECTIVE OPTIMIZATION Petri Eskelinen 1, Kaisa Miettinen 2 Internatonal Conference 20th EURO Mn Conference Contnuous Optmaton and Knowledge-Based Technologes (EurOPT-2008) May 20 23, 2008, Nernga, LITHUANIA ISBN 978-9955-28-283-9 L. Saalausas, G.W. Weber and E.

More information

Reduce Distillation Column Cost by Hybrid Particle Swarm and Ant

Reduce Distillation Column Cost by Hybrid Particle Swarm and Ant From the SelectedWorks of Dr. Sandp Kumar Lahr Summer July 20, 2016 Reduce Dstllaton Column Cost by Hybrd Partcle Swarm and Ant Dr. Sandp k lahr chnmaya lenka Avalable at: https://works.bepress.com/sandp_lahr/33/

More information

Modeling Form for On-line Following of Musical Performances

Modeling Form for On-line Following of Musical Performances Modelng Form for On-lne Followng of Muscal Performances Bryan Pardo 1 and Wllam Brmngham 2 1 Computer Scence Department, Northwestern Unversty, 1890 Maple Ave, Evanston, IL 60201 2 Department of Math and

More information

Integration of Internet of Thing Technology in Digital Energy Network with Dispersed Generation

Integration of Internet of Thing Technology in Digital Energy Network with Dispersed Generation Amercan Scentfc Research Journal for Engneerng, Technology, and Scences (ASRJETS) ISS (Prnt) 2313-4410, ISS (Onlne) 2313-4402 Global Socety of Scentfc Research and Researchers http://asrjetsjournal.org/

More information

THE IMPORTANCE OF ARM-SWING DURING FORWARD DIVE AND REVERSE DIVE ON SPRINGBOARD

THE IMPORTANCE OF ARM-SWING DURING FORWARD DIVE AND REVERSE DIVE ON SPRINGBOARD THE MPORTANCE OF ARM-SWNG DURNG FORWARD DVE AND REVERSE DVE ON SPRNGBOARD Ken Yokoyama Laboratory of Bomechancs Faculty ofeducaton Kanazawa Unversty Kanazawa, Japan J unjro Nagano Department of Physcal

More information

Study on the location of building evacuation indicators based on eye tracking

Study on the location of building evacuation indicators based on eye tracking Study on the locaton of buldng evacuaton ndcators based on eye trackng Yue L Tsnghua Unversty yue-l5@malstsnghuaeducn Png hang Tsnghua Unversty zhangp@malstsnghuaeducn Hu hang Tsnghua Unversty, zhhu@tsnghuaeducn

More information

MODELING AND ANALYZING THE VOCAL TRACT UNDER NORMAL AND STRESSFUL TALKING CONDITIONS

MODELING AND ANALYZING THE VOCAL TRACT UNDER NORMAL AND STRESSFUL TALKING CONDITIONS MODELING AND ANALYZING THE VOCAL TRACT UNDER NORMAL AND STRESSFUL TALING CONDITIONS Ismal Shahn and Naeh Botros 2 Electrcal/Electroncs and Comuter Engneerng Deartment Unversty of Sharjah, P. O. Box 27272,

More information

Correcting Image Placement Errors Using Registration Control (RegC ) Technology In The Photomask Periphery

Correcting Image Placement Errors Using Registration Control (RegC ) Technology In The Photomask Periphery Correctng Image Placement Errors Usng Regstraton Control (RegC ) Technology In The Photomask Perphery Av Cohen 1, Falk Lange 2 Guy Ben-Zv 1, Erez Gratzer 1, Dmtrev Vladmr 1 1. Carl Zess SMS Ltd., Karmel

More information

Failure Rate Analysis of Power Circuit Breaker in High Voltage Substation

Failure Rate Analysis of Power Circuit Breaker in High Voltage Substation T. Suwanasr, M. T. Hlang and C. Suwanasr / GMSAR Internatonal Journal 8 (2014) 1-6 Falure Rate Analyss of Power Crcut Breaker n Hgh Voltage Substaton Thanapong Suwanasr, May Thandar Hlang and Cattareeya

More information

AMP-LATCH* Ultra Novo mm [.025 in.] Ribbon Cable 02 MAR 12 Rev C

AMP-LATCH* Ultra Novo mm [.025 in.] Ribbon Cable 02 MAR 12 Rev C AMP-LATCH* Ultra Novo Applcaton Specfcaton Receptacle Connectors for 114-40056 0.64 mm [.025 n.] Rbbon Cable 02 MAR 12 All numercal values are n metrc unts [wth U.S. customary unts n brackets]. Dmensons

More information

AN INTERACTIVE APPROACH FOR MULTI-CRITERIA SORTING PROBLEMS

AN INTERACTIVE APPROACH FOR MULTI-CRITERIA SORTING PROBLEMS AN INTERACTIVE APPROACH FOR MULTI-CRITERIA SORTING PROBLEMS A THESIS SUBMITTED TO THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES OF MIDDLE EAST TECHNICAL UNIVERSITY BY BURAK KESER IN PARTIAL FULFILLMENT

More information

Simple VBR Harmonic Broadcasting (SVHB)

Simple VBR Harmonic Broadcasting (SVHB) mple VBR Harmonc Broadcastng (VHB) Hsang-Fu Yu ab, Hung-hang Yang a, Y-Mng hen c, -Mng Tseng a, and hen-y Kuo a a Dep. of omputer cence & Informaton Engneerng, atonal entral Unversty, Tawan b omputer enter,

More information

Quantization of Three-Bit Logic for LDPC Decoding

Quantization of Three-Bit Logic for LDPC Decoding Proceedngs of the World Congress on Engneerng and Computer Scence 2011 Vol II, October 19-21, 2011, San Francsco, USA Quantzaton of Three-Bt Logc for LDPC Decodng Raymond Moberly and Mchael E. O'Sullvan

More information

RIAM Local Centre Woodwind, Brass & Percussion Syllabus

RIAM Local Centre Woodwind, Brass & Percussion Syllabus 8 RIAM Local Centre Woodwnd, Brass & Percusson Syllabus 2015-2018 AURAL REQUIREMENTS AND THEORETICAL QUESTIONS REVISED FOR ALL PRACTICAL SUBJECTS AURAL TESTS From Elementary to Grade V ths area s worth

More information

Production of Natural Penicillins by Strains of Penicillium chrysogenutn

Production of Natural Penicillins by Strains of Penicillium chrysogenutn Producton of Natural Pencllns by Strans of Pencllum chrysogenutn a J. FUSK and ЬЕ. WELWRDOVÁ ^Department of Mcrobology and Bochemstry, Slovak Techncal Unversty, Bratslava b Botka, Slovenská Ľupča Receved

More information

Craig Webre, Sheriff Personnel Division/Law Enforcement Complex 1300 Lynn Street Thibodaux, Louisiana 70301

Craig Webre, Sheriff Personnel Division/Law Enforcement Complex 1300 Lynn Street Thibodaux, Louisiana 70301 DATE OF APPLCATON: Craig Webre, Sheriff Personnel Division/Law Enforcement Complex 1300 Lynn Street Thibodaux, Louisiana 70301 N GENERAL EMAL ADDRESS: For Local Calls - (985) 532-4380 (985) 446-2255 (985)

More information

Detecting Errors in Blood-Gas Measurement by Analysiswith Two Instruments

Detecting Errors in Blood-Gas Measurement by Analysiswith Two Instruments CLIN. CHEM. 33/4, 512-517 (1987) Detectng Errors n Blood-Gas Measurement by Analysswth Two Instruments LouIs F. Metzger, Wllam B. Stauffer, Ann V. Kruplnskl, Rchard P. MIIlman,3 George S. Cembrowskl,2

More information

Automated composer recognition for multi-voice piano compositions using rhythmic features, n-grams and modified cortical algorithms

Automated composer recognition for multi-voice piano compositions using rhythmic features, n-grams and modified cortical algorithms Complex Intell. Syst. (2018) 4:55 65 https://do.org/10.1007/s40747-017-0052-x ORIGINAL ARTICLE Automated composer recognton for mult-voce pano compostons usng rhythmc features, n-grams and modfed cortcal

More information

Multi-Line Acquisition With Minimum Variance Beamforming in Medical Ultrasound Imaging

Multi-Line Acquisition With Minimum Variance Beamforming in Medical Ultrasound Imaging IEEE Transactons on Ultrasoncs, Ferroelectrcs, and Frequency Control, vol. 60, no. 12, Decemer 2013 2521 Mult-Lne Acquston Wth Mnmum Varance Beamformng n Medcal Ultrasound Imagng Ad Ranovch, Zv Fredman,

More information

Critical Path Reduction of Distributed Arithmetic Based FIR Filter

Critical Path Reduction of Distributed Arithmetic Based FIR Filter Crtcal Path Reducton of strbuted rthmetc Based FIR Flter Sunta Badave epartment of Electrcal and Electroncs Engneerng.I.T, urangabad aharashtra, Inda njal Bhalchandra epartment of Electroncs and Telecommuncaton

More information

Improving Reliability and Energy Efficiency of Disk Systems via Utilization Control

Improving Reliability and Energy Efficiency of Disk Systems via Utilization Control Ths paper appeared n the Proceedngs of the 2th IEEE Symposum on Computers and Communcatons (ISCC'08, Marrakech, Morocco, July 2008. Improvng Relablty and Energy Effcency of Dsk Systems va Utlzaton Control

More information

A STUDY OF TRUMPET ENVELOPES

A STUDY OF TRUMPET ENVELOPES A STUDY OF TRUMPET ENVELOPES Roger B. Dannenberg, Hank Pellern, and Istvan Dereny School of Computer Scence, Carnege Mellon Unversty Pttsburgh, PA 15213 USA rbd@cs.cmu.edu, hank.pellern@andrew.cmu.edu,

More information

User s manual. Digital control relay SVA

User s manual. Digital control relay SVA User s manual Dgtal control relay DISIBEINT ELECTRONIC S.L, has been present n the feld of the manufacture of components for the ndustral automaton for more than years, and mantans n constant evoluton

More information

3 Part differentiation, 20 parameters, 3 histograms Up to patient results (including histograms) can be stored

3 Part differentiation, 20 parameters, 3 histograms Up to patient results (including histograms) can be stored st Techncal Specfcatons Desgned n France Wth a rch past and a professonal experence bult-up over 35 years, SFRI s a French nvatve company commtted to developng preon In Vtro Dst solutons. SFRI has bult

More information

Analysis of Subscription Demand for Pay-TV

Analysis of Subscription Demand for Pay-TV Analyss of Subscrpton Demand for Pay-TV Manabu Shshkura Researcher Insttute for Informaton and Communcatons Polcy 2-1-2 Kasumgasek, Chyoda-ku Tokyo 110-8926 Japan m-shshkura@soumu.go.jp Tel: 03-5253-5496

More information

Lost on the Web: Does Web Distribution Stimulate or Depress Television Viewing?

Lost on the Web: Does Web Distribution Stimulate or Depress Television Viewing? Lost on the Web: Does Web Dstrbuton Stmulate or Depress Televson Vewng? Joel Waldfogel The Wharton School Unversty of Pennsylvana August 10, 2007 Prelmnary comments welcome Abstract In the past few years,

More information

AIAA Optimal Sampling Techniques for Zone- Based Probabilistic Fatigue Life Prediction

AIAA Optimal Sampling Techniques for Zone- Based Probabilistic Fatigue Life Prediction AIAA 2002-383 Optmal Samplng Technques or Zone- Based Probablstc Fatgue Le Predcton M. P. Enrght Southwest Research Insttute San Antono, TX H. R. Mllwater Unversty o Texas at San Antono San Antono, TX

More information

T541 Flat Panel Monitor User Guide ENGLISH

T541 Flat Panel Monitor User Guide ENGLISH T541 Flat Panel Montor User Gude ENGLISH Frst Edton (June / 2002) Note : For mportant nformaton, refer to the Montor Safety and Warranty manual that comes wth ths montor. Ths publcaton could contan techncal

More information

Modular Plug Connectors (Standard and Small Conductor)

Modular Plug Connectors (Standard and Small Conductor) Modular Plug Connectors (Standard and Small Conductor) Applcaton Specfcaton 114-6016 04 APR 11 All numercal values are n metrc unts [wth U.S. customary unts n brackets]. Dmensons are n mllmeters [and nches].

More information

Scalable QoS-Aware Disk-Scheduling

Scalable QoS-Aware Disk-Scheduling Scalable QoS-Aware Dsk-Schedulng Wald G. Aref Khaled El-Bassyoun Ibrahm Kamel Mohamed F. Mokbel Department of Computer Scences, urdue Unversty, West Lafayette, IN 47907-1398 anasonc Informaton and Networkng

More information

Color Monitor. L200p. English. User s Guide

Color Monitor. L200p. English. User s Guide Color Montor L200p User s Gude Englsh Frst Edton (February / 2003) Note : For mportant nformaton, refer to the Montor Safety and Warranty manual that comes wth ths montor. Contents ENGLISH Safety (Read

More information

SKEW DETECTION AND COMPENSATION FOR INTERNET AUDIO APPLICATIONS. Orion Hodson, Colin Perkins, and Vicky Hardman

SKEW DETECTION AND COMPENSATION FOR INTERNET AUDIO APPLICATIONS. Orion Hodson, Colin Perkins, and Vicky Hardman SKEW DETECTION AND COMPENSATION FOR INTERNET AUDIO APPLICATIONS Oron Hodson, Coln Perkns, and Vcky Hardman Department of Computer Scence Unversty College London Gower Street, London, WC1E 6BT, UK. ABSTRACT

More information

Product Information. Manual change system HWS

Product Information. Manual change system HWS Product Informaton HWS HWS Flexble. Compact. Productve. HWS manual change system Manual tool change system wth ntegrated ar feed-through and optonal electrc feed-through Feld of applcaton Excellently sutable

More information

Small Area Co-Modeling of Point Estimates and Their Variances for Domains in the Current Employment Statistics Survey

Small Area Co-Modeling of Point Estimates and Their Variances for Domains in the Current Employment Statistics Survey Small Area Co-Modelng of Pont Estmates and Ther Varances for Domans n the Current Employment Statstcs Survey Jule Gershunskaya, Terrance D. Savtsky U.S. Bureau of Labor Statstcs FCSM, March 2018 Dsclamer:

More information

Product Information. Manual change system HWS

Product Information. Manual change system HWS Product Informaton HWS HWS Flexble. Compact. Productve. HWS manual change system Manual tool change system wth ntegrated ar feed-through and optonal electrc feed-through Feld of applcaton Excellently sutable

More information

Fast Intra-Prediction Mode Decision in H.264/AVC Based on Macroblock Properties

Fast Intra-Prediction Mode Decision in H.264/AVC Based on Macroblock Properties Fast Intra-Predcton Mode Decson n H.264/AVC Based on Macroblock Propertes Abstract Intra-predcton s a wdely used tecnque n ntra codng. H.264/AVC adopts rate-dstorton optmzaton (RDO) tecnque to obtan te

More information

S Micro--Strip Tool in. S Combination Strip Tool ( ) S Cable Holder Assembly (Used only

S Micro--Strip Tool in. S Combination Strip Tool ( ) S Cable Holder Assembly (Used only Instructon Sheet LghtCrmp* Plus LC 408-10103 (for Jacketed Cable) Connectors 18 AUG 09 Rear Protectve Cap Termnaton CoverG Boot Connector Assembly Crmp Eyelet Duplex Clp G Connector kt s shpped wth these

More information

Product Bulletin 40C 40C-10R 40C-20R 40C-114R. Product Description For Solvent, Eco-Solvent, UV and Latex Inkjet and Screen Printing 3-mil vinyl films

Product Bulletin 40C 40C-10R 40C-20R 40C-114R. Product Description For Solvent, Eco-Solvent, UV and Latex Inkjet and Screen Printing 3-mil vinyl films Product Bulletn 40C Revson D, Effectve February 2016 (Replaces C, Apr. 15) 40C-10R 40C-20R 40C-114R Product Descrpton For Solvent, Eco-Solvent, UV and Latex Inkjet and Screen Prntng 3-ml vnyl flms Quck

More information

Product Information. Miniature rotary unit ERD

Product Information. Miniature rotary unit ERD Product Informaton ERD ERD Fast. Compact. Flexble. ERD torque motor Powerful torque motor wth absolute encoder and electrc and pneumatc rotary feed-through Feld of applcaton For all applcatons wth exceptonal

More information

Simple Solution for Designing the Piecewise Linear Scalar Companding Quantizer for Gaussian Source

Simple Solution for Designing the Piecewise Linear Scalar Companding Quantizer for Gaussian Source 94 J. NIKOIĆ, Z. PERIĆ,. VEIMIROVIĆ, SIMPE SOUTION FOR DESIGNING THE PIECEWISE INEAR SCAAR Smle Soluton for Desgnng the Pecewse near Scalar Comandng Quantzer for Gaussan Source Jelena NIKOIĆ, Zoran PERIĆ,

More information

Clock Synchronization in Satellite, Terrestrial and IP Set-top Box for Digital Television

Clock Synchronization in Satellite, Terrestrial and IP Set-top Box for Digital Television Clock Synchronzaton n Satellte, Terrestral and IP Set-top Box for Dgtal Televson THESIS Submtted n partal fulflment of the requrements for the degree of DOCTOR OF PHILOSOPHY by MONIKA JAIN Under the Supervson

More information

CASH TRANSFER PROGRAMS WITH INCOME MULTIPLIERS: PROCAMPO IN MEXICO

CASH TRANSFER PROGRAMS WITH INCOME MULTIPLIERS: PROCAMPO IN MEXICO FCND DP No. 99 FCND DISCUSSION PAPER NO. 99 CASH TRANSFER PROGRAMS WITH INCOME MULTIPLIERS: PROCAMPO IN MEXICO Elsabeth Sadoulet, Alan de Janvry, and Benjamn Davs Food Consumpton and Nutrton Dvson Internatonal

More information

User Manual. AV Router. High quality VGA RGBHV matrix that distributes signals directly. Controlled via computer.

User Manual. AV Router. High quality VGA RGBHV matrix that distributes signals directly. Controlled via computer. User Manual AV Router Hgh qualty VGA RGBHV matrx that dstrbutes sgnals drectly. Controlled va computer. Notce: : The nmaton contaned n ths document s subject to change wthout notce. SmartAVI makes no warranty

More information

INSTRUCTION MANUAL FOR THE INSTALLATION, USE AND MAINTENANCE OF THE REGULATOR GENIUS POWER COMBI

INSTRUCTION MANUAL FOR THE INSTALLATION, USE AND MAINTENANCE OF THE REGULATOR GENIUS POWER COMBI NSTRUCTON MANUAL FOR THE NSTALLATON, USE AND MANTENANCE OF THE REGULATOR GENUS POWER COMB (TRANSLATON OF THE ORGNAL NSTRUCTON MANUAL N TALAN) PRELMNARY VERSON WARRANTY The devce s guaranteed 24 months

More information

Loewe bild 7.65 OLED. Set-up options. Loewe bild 7 cover Incl. Back cover. Loewe bild 7 cover kit Incl. Back cover and Speaker cover

Loewe bild 7.65 OLED. Set-up options. Loewe bild 7 cover Incl. Back cover. Loewe bild 7 cover kit Incl. Back cover and Speaker cover Product nformaton Loewe bld 7.65 Page of March 07 Loewe bld 7.65 OLED EU energy effcency class: B Screen dagonal (n cm) / Screen dagonal (n nch): 64 / 65 Power consumpton ON (n W): 80 Annual energy consumpton

More information

FPGA Implementation of Cellular Automata Based Stream Cipher: YUGAM-128

FPGA Implementation of Cellular Automata Based Stream Cipher: YUGAM-128 ISSN (Prnt) : 2320 3765 ISSN (Onlne): 2278 8875 Internatonal Journal of Advanced Research n Electrcal, Electroncs and Instrumentaton Engneerng An ISO 3297: 2007 Certfed Organzaton Vol. 3, Specal Issue

More information

9! VERY LARGE IN THEIR CONCERNS. AND THEREFORE, UH, i

9! VERY LARGE IN THEIR CONCERNS. AND THEREFORE, UH, i 340 WELL, alack PAJAMAS WAS A SOMEWHAT METAPHORCAL 2 TERM. MANY VETNAMESE PEASANTS TENDED TO WEAR 3 BLACK PAJAMAS, BUT WHAT AM REFERRNG TO S THAT 4 OUTSDE OF THE NORTH VETNAMESE UNTS AND ~OME OF 5 THE

More information

Environmental Reviews. Cause-effect analysis for sustainable development policy

Environmental Reviews. Cause-effect analysis for sustainable development policy Envronmental Revews Cause-effect analyss for sustanable development polcy Journal: Envronmental Revews Manuscrpt ID er-2016-0109.r2 Manuscrpt Type: Revew Date Submtted by the Author: 24-Feb-2017 Complete

More information

Loewe bild 5.55 oled. Modular Design Flexible configuration with individual components. Set-up options. TV Monitor

Loewe bild 5.55 oled. Modular Design Flexible configuration with individual components. Set-up options. TV Monitor Product nformaton Loewe bld 5.55 oled Page of 3 Loewe bld 5.55 oled EU energy effcency class: B Screen dagonal (n cm) / Screen dagonal (n nch): 39 / 55 Power consumpton ON (n W): 50 Annual energy consumpton

More information

DT-500 OPERATION MANUAL MODE D'EMPLOI MANUAL DE MANEJO MANUAL DE OPERA(_._,O. H.-,lri-D PROJECTOR PROJECTEUR PROYECTOR PROJETOR

DT-500 OPERATION MANUAL MODE D'EMPLOI MANUAL DE MANEJO MANUAL DE OPERA(_._,O. H.-,lri-D PROJECTOR PROJECTEUR PROYECTOR PROJETOR TM PROJECTOR PROJECTEUR PROYECTOR PROJETOR DT-500 OPERATION MANUAL MODE D'EMPLOI MANUAL DE MANEJO MANUAL DE OPERA(_._,O 8 f f 8 H.-,lr-D _I_H DEFINmON_TIM_IA I_T_RFACE Before usng the projector, please

More information

Joint Image and Text Representation for Aesthetics Analysis

Joint Image and Text Representation for Aesthetics Analysis Joint Image and Text Representation for Aesthetics Analysis Ye Zhou 1, Xin Lu 2, Junping Zhang 1, James Z. Wang 3 1 Fudan University, China 2 Adobe Systems Inc., USA 3 The Pennsylvania State University,

More information

Sealed Circular LC Connector System Plug

Sealed Circular LC Connector System Plug Sealed Crcular LC Connector System Plug Instructon Sheet Kt 1828618- [ ], Receptacle Kt 1828619- [ ], 408-10079 and EMI Receptacle Kt 1985193- [ ] 07 APR 11 Plug Kt 1828618 -[ ] Cable Fttng Receptacle

More information

Printer Specifications

Printer Specifications : Characterfonts:! Fort Pont 7P 0.5 pl Ptch 5cpl, Ocpl, 2cpl Proptlonel Epson Draf!o 0 lo j Epson Cower 0 0 0 Epson Roman O O O 0 Epson San6 Sent 0 O O O Epson Presllge j0 0 ~Epson Scnpt O 0 Epson Sormt

More information

Discussion Paper Series

Discussion Paper Series Doshsha Unversty Center for the Study of the Creatve Economy Dscusson Paper Seres No. 2013-04 Nonlnear Effects of Superstar Collaboraton: Why the Beatles Succeeded but Broke Up Tadash Yag Dscusson Paper

More information

IMAGE AESTHETIC PREDICTORS BASED ON WEIGHTED CNNS. Oce Print Logic Technologies, Creteil, France

IMAGE AESTHETIC PREDICTORS BASED ON WEIGHTED CNNS. Oce Print Logic Technologies, Creteil, France IMAGE AESTHETIC PREDICTORS BASED ON WEIGHTED CNNS Bin Jin, Maria V. Ortiz Segovia2 and Sabine Su sstrunk EPFL, Lausanne, Switzerland; 2 Oce Print Logic Technologies, Creteil, France ABSTRACT Convolutional

More information

IN DESCRIBING the tape transport of

IN DESCRIBING the tape transport of Apparatus For Magnetc Storage on Three-Inch Wde Tapes R. B. LAWRANCE R. E. WILKINS R. A. PENDLETON IN DESCRIBING the tape transport of the DATAmatc 1, t s perhaps well to begn by revewng the nfluental

More information

(12) Ulllted States Patent (10) Patent N0.: US 8,269,970 B2 P0lid0r et a]. (45) Date of Patent: Sep. 18, 2012

(12) Ulllted States Patent (10) Patent N0.: US 8,269,970 B2 P0lid0r et a]. (45) Date of Patent: Sep. 18, 2012 US008269970B2 (12) Ulllted States Patent (10) Patent N0.: P0ld0r et a]. (45) Date of Patent: Sep. 18, 12 (54) OPTICAL COMPARATOR WITH DIGITAL 6,945,652 B2 9/05 sakqta et a1 GAGE 7,058,109 B2* 6/06 Davs.....

More information

Product Information. Universal swivel units SRU-plus

Product Information. Universal swivel units SRU-plus Product Informaton Unversal swvel unts SRU-plus SRU-plus Unversal swvel unts Robust. Fast. Hgh Performance. SRU-plus unversal rotary actuator Unversal unt for pneumatc swvel and turnng movements. Feld

More information

in Partial For the Degree of

in Partial For the Degree of 37q h8( sta. co AN ANALYSS OF ROBERT NATHANEL DETT'S N THE BOTTOMS THESS Presented to the Graduate ouncl of the North Texas State Unversty n Partal Fulfllment of the Requrements For the Degree of MASTER

More information

Image Restoration using Multilayer Neural Networks with Minimization of Total Variation Approach

Image Restoration using Multilayer Neural Networks with Minimization of Total Variation Approach IJCSI Internatonal Jornal of Compter Scence Isses, Vol., Isse, No, Janar 4 ISSN (Prnt): 694-84 ISSN (Onlne): 694-784 www.ijcsi.org 6 Image Restoraton sng Mltlaer Neral Networs wth Mnmzaton of Total Varaton

More information

User guide. Receiver-In-Ear hearing aids. resound.com

User guide. Receiver-In-Ear hearing aids. resound.com User gude Recever-In-Ear hearng ads resound.com 400786011US-17.07-Rev.A.ndd 1 20-07-2017 12:52:40 Left Hearng Ad Rght Hearng Ad Seral number Seral number Model number Model number Recever type Recever

More information

Conettix D6600/D6100IPv6 Communications Receiver/Gateway Quick Start

Conettix D6600/D6100IPv6 Communications Receiver/Gateway Quick Start Conettx / Communcatons Recever/Gateway Quck Start.0 Parts Lst able : Conettx System Components Qty. Descrpton Conettx Communcatons Recever/Gateway AC power cord Battery cable P660 I/O cable P660 Rack mount

More information

SWS 160. Moment loading. Technical data. M x max Nm M y max Nm. M z max Nm

SWS 160. Moment loading. Technical data. M x max Nm M y max Nm. M z max Nm Moment loadng M x max. 7170 Nm M y max. 7170 Nm M z max. 3800 Nm Ths s the max. sum of all forces and moments (from acceleraton forces and moments, process forces or moments, emergency stop stuatons, etc.)

More information

INTERCOM SMART VIDEO DOORBELL. Installation & Configuration Guide

INTERCOM SMART VIDEO DOORBELL. Installation & Configuration Guide INTERCOM SMART VIDEO DOORBELL Installaton & Confguraton Gude ! Important safety nformaton Read ths manual before attemptng to nstall the devce! Falure to observe recommendatons ncluded n ths manual may

More information

Deep Aesthetic Quality Assessment with Semantic Information

Deep Aesthetic Quality Assessment with Semantic Information 1 Deep Aesthetic Quality Assessment with Semantic Information Yueying Kao, Ran He, Kaiqi Huang arxiv:1604.04970v3 [cs.cv] 21 Oct 2016 Abstract Human beings often assess the aesthetic quality of an image

More information

JTAG / Boundary Scan. Multidimensional JTAG / Boundary Scan Instrumentation. Get the total Coverage!

JTAG / Boundary Scan. Multidimensional JTAG / Boundary Scan Instrumentation. Get the total Coverage! JTAG / Boundary Scan Multdmensonal JTAG / Boundary Scan Instrumentaton IEEE 1149.6 IEEE 1149.1 IEEE 1149.7 Multdmensonal JTAG / Boundary Scan Instrumentaton IEEE 1149.4 IEEE 1532 Get the total Coverage!

More information

Expressive Musical Timing

Expressive Musical Timing Axel Berndt, Tlo Hähnel Department of Smulaton and Graphcs Otto-von-Guercke Unversty of Magdeburg {aberndt tlo}@sg.cs.un-magdeburg.de Abstract. Tmng s crucal for the qualty of expressve musc performances.

More information

zenith Installation and Operating Guide HodelNumber I Z42PQ20 [ PLASHATV

zenith Installation and Operating Guide HodelNumber I Z42PQ20 [ PLASHATV Installaton and Operatng Gude HodelNumber I Z42PQ20 PLASHATV To vew the extended verson of owner's manual that contans the advanced features of ths TV set, vst our webste at http://www.enthservce.com Ths

More information

Bachelor s Degree Programme (BDP)

Bachelor s Degree Programme (BDP) EEG-01/ BEGE-101 Bachelor s Degree Programme (BDP) ASSIGNMENT (for July 2018 and January 2019 Sessons) EEG-01/BEGE-101 ELECTIVE COURSE IN ENGLISH School of Humantes Indra Gandh Natonal Open Unersty Madan

More information

A question of character. Loewe Connect ID.

A question of character. Loewe Connect ID. A queston of character. Loewe Connect ID. Modern. Etquette you can learn, character s nnate. You make a clear dstncton between good manners and genune style. Your s s lke yourself: unfussy, busnesslke

More information

CS 1674: Intro to Computer Vision. Intro to Recognition. Prof. Adriana Kovashka University of Pittsburgh October 24, 2016

CS 1674: Intro to Computer Vision. Intro to Recognition. Prof. Adriana Kovashka University of Pittsburgh October 24, 2016 CS 1674: Intro to Computer Vision Intro to Recognition Prof. Adriana Kovashka University of Pittsburgh October 24, 2016 Plan for today Examples of visual recognition problems What should we recognize?

More information

DeepID: Deep Learning for Face Recognition. Department of Electronic Engineering,

DeepID: Deep Learning for Face Recognition. Department of Electronic Engineering, DeepID: Deep Learning for Face Recognition Xiaogang Wang Department of Electronic Engineering, The Chinese University i of Hong Kong Machine Learning with Big Data Machine learning with small data: overfitting,

More information

Predicting Aesthetic Radar Map Using a Hierarchical Multi-task Network

Predicting Aesthetic Radar Map Using a Hierarchical Multi-task Network Predicting Aesthetic Radar Map Using a Hierarchical Multi-task Network Xin Jin 1,2,LeWu 1, Xinghui Zhou 1, Geng Zhao 1, Xiaokun Zhang 1, Xiaodong Li 1, and Shiming Ge 3(B) 1 Department of Cyber Security,

More information

US Al (19) United States (12) Patent Application Publication (10) Pub. No.: US 2014/ A1 ABE (43) Pub. Date: Jun.

US Al (19) United States (12) Patent Application Publication (10) Pub. No.: US 2014/ A1 ABE (43) Pub. Date: Jun. . US 20140178045Al (19) United States (12) Patent Application Publication (10) Pub. No.: US 2014/0178045 A1 ABE (43) Pub. Date: Jun. 26, 2014 (54) VDEO PLAYBACK DEVCE, VDEO Publication Classi?cation PLAYBACK

More information

www. ElectricalPartManuals. com l Basler Electric VOLTAGE REGULATOR FEATURES: CLASS 300 EQUIPMENT AVC63 4 FEATURES AND APPLICATIONS

www. ElectricalPartManuals. com l Basler Electric VOLTAGE REGULATOR FEATURES: CLASS 300 EQUIPMENT AVC63 4 FEATURES AND APPLICATIONS Using enhanced technology, the AVC63-4 voltage regulator is designed for use on 50/60 Hz brushless generators. This encapsulated regulator is economical, small in size, ruggedly constructed, and incorporates

More information