%PDF-1.3 1 0 obj << /Kids [ 4 0 R 5 0 R 6 0 R 7 0 R 8 0 R 9 0 R 10 0 R 11 0 R 12 0 R 13 0 R 14 0 R 15 0 R 16 0 R ] /Type /Pages /Count 13 >> endobj 2 0 obj << /Subject (Neural Information Processing Systems http\072\057\057nips\056cc\057) /Publisher (Curran Associates\054 Inc\056) /Language (en\055US) /Created (2019) /EventType (Poster) /Description-Abstract (We present a simple\054 flexible\054 and general framework titled Partial Registration Network \050PRNet\051\054 for partial\055to\055partial point cloud registration\056 Inspired by recently\055proposed learning\055based methods for registration\054 we use deep networks to tackle non\055convexity of the alignment and partial correspondence problem\056 While previous learning\055based methods assume the entire shape is visible\054 PRNet is suitable for partial\055to\055partial registration\054 outperforming PointNetLK\054 DCP\054 and non\055learning methods on synthetic data\056 PRNet is self\055supervised\054 jointly learning an appropriate geometric representation\054 a keypoint detector that finds points in common between partial views\054 and keypoint\055to\055keypoint correspondences\056 We show PRNet predicts keypoints and correspondences consistently across views and objects\056 Furthermore\054 the learned representation is transferable to classification\056) /Producer (PyPDF2) /Title (PRNet\072 Self\055Supervised Learning for Partial\055to\055Partial Registration) /Date (2019) /ModDate (D\07220200213022358\05508\04700\047) /Published (2019) /Type (Conference Proceedings) /firstpage (8814) /Book (Advances in Neural Information Processing Systems 32) /Description (Paper accepted and presented at the Neural Information Processing Systems Conference \050http\072\057\057nips\056cc\057\051) /Editors (H\056 Wallach and H\056 Larochelle and A\056 Beygelzimer and F\056 d\047Alch\351\055Buc and E\056 Fox and R\056 Garnett) /Author (Yue Wang\054 Justin M\056 Solomon) /lastpage (8826) >> endobj 3 0 obj << /Type /Catalog /Pages 1 0 R >> endobj 4 0 obj << /Parent 1 0 R /Contents 17 0 R /Resources 18 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 43 0 R /Type /Page >> endobj 5 0 obj << /Parent 1 0 R /Contents 66 0 R /Resources 67 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 99 0 R /Type /Page >> endobj 6 0 obj << /Parent 1 0 R /Contents 325 0 R /Resources 326 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 375 0 R /Type /Page >> endobj 7 0 obj << /Parent 1 0 R /Contents 481 0 R /Resources 482 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 502 0 R /Type /Page >> endobj 8 0 obj << /Parent 1 0 R /Contents 543 0 R /Resources 544 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 573 0 R /Type /Page >> endobj 9 0 obj << /Parent 1 0 R /Contents 639 0 R /Resources 640 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 643 0 R /Type /Page >> endobj 10 0 obj << /Parent 1 0 R /Contents 729 0 R /Resources 730 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 733 0 R /Type /Page >> endobj 11 0 obj << /Parent 1 0 R /Contents 879 0 R /Resources 880 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 887 0 R /Type /Page >> endobj 12 0 obj << /Parent 1 0 R /Contents 908 0 R /Resources 909 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Annots 912 0 R /Type /Page >> endobj 13 0 obj << /Parent 1 0 R /Contents 938 0 R /Resources 939 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Type /Page >> endobj 14 0 obj << /Parent 1 0 R /Contents 940 0 R /Resources 941 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Type /Page >> endobj 15 0 obj << /Parent 1 0 R /Contents 942 0 R /Resources 943 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Type /Page >> endobj 16 0 obj << /Parent 1 0 R /Contents 951 0 R /Resources 952 0 R /Rotate 0 /MediaBox [ 0 0 612 792 ] /Type /Page >> endobj 17 0 obj << /Length 4142 /Filter /FlateDecode >> stream xZێ}WD "ۗN[8M3hPyP)W+]ksx5Mm_6t~tjRg륛ԏ0Kp1?9vq$mz/Idn8\/}ww7:=ǽ-T78q?Vr Ƿeq~YiÍ7Ο^JowUwq ح|&8nY<>߬&_ChJ$Yĩ!ԛ}j>Xur߬ރCI%,L6x6ը74=Hǩ J#Fgfg˳̊l+xM^|h!4-uȣ~uO/:e6 (kn$qb?8@M2q)U'W-eQB\O6h矪{ۛP B,&?%ā'٦ >(p*m{)o˛ϛZpZˇfABmFY0!6>ԲN#SS)84Hۅ>( ʽAܡ-<A!kr0#ESۜ8 &Hseov .!M\B I'$@]Y,l6_]cvy+'m09UJФK^DT4LTuSk 2tnץHŧb2V>VM95ap