%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 ] /Type /Pages /Count 9 >> 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 (2012) /Description-Abstract (We trained a large\054 deep convolutional neural network to classify the 1\0563 million high\055resolution images in the LSVRC\0552010 ImageNet training set into the 1000 different classes\056 On the test data\054 we achieved top\0551 and top\0555 error rates of 39\0567\134\045 and 18\0569\134\045 which is considerably better than the previous state\055of\055the\055art results\056 The neural network\054 which has 60 million parameters and 500\054000 neurons\054 consists of five convolutional layers\054 some of which are followed by max\055pooling layers\054 and two globally connected layers with a final 1000\055way softmax\056 To make training faster\054 we used non\055saturating neurons and a very efficient GPU implementation of convolutional nets\056 To reduce overfitting in the globally connected layers we employed a new regularization method that proved to be very effective\056) /Producer (Python PDF Library \055 http\072\057\057pybrary\056net\057pyPdf\057) /Title (ImageNet Classification with Deep Convolutional Neural Networks) /Date (2012) /ModDate (D\07220140423102144\05507\04700\047) /Published (2012) /Type (Conference Proceedings) /firstpage (1097) /Book (Advances in Neural Information Processing Systems 25) /Description (Paper accepted and presented at the Neural Information Processing Systems Conference \050http\072\057\057nips\056cc\057\051) /Editors (F\056 Pereira and C\056J\056C\056 Burges and L\056 Bottou and K\056Q\056 Weinberger) /Author (Alex Krizhevsky\054 Ilya Sutskever\054 Geoffrey E\056 Hinton) /lastpage (1105) >> endobj 3 0 obj << /Type /Catalog /Pages 1 0 R >> endobj 4 0 obj << /Contents 13 0 R /Parent 1 0 R /Type /Page /Resources 14 0 R /MediaBox [ 0 0 612 792 ] >> endobj 5 0 obj << /Contents 28 0 R /Parent 1 0 R /Type /Page /Resources 29 0 R /MediaBox [ 0 0 612 792 ] >> endobj 6 0 obj << /Contents 38 0 R /Parent 1 0 R /Type /Page /Resources 39 0 R /MediaBox [ 0 0 612 792 ] >> endobj 7 0 obj << /Contents 65 0 R /Parent 1 0 R /Type /Page /Resources 66 0 R /MediaBox [ 0 0 612 792 ] >> endobj 8 0 obj << /Contents 71 0 R /Parent 1 0 R /Type /Page /Resources 72 0 R /MediaBox [ 0 0 612 792 ] >> endobj 9 0 obj << /Contents 80 0 R /Parent 1 0 R /Type /Page /Resources 81 0 R /MediaBox [ 0 0 612 792 ] >> endobj 10 0 obj << /Contents 94 0 R /Parent 1 0 R /Type /Page /Resources 95 0 R /MediaBox [ 0 0 612 792 ] >> endobj 11 0 obj << /Contents 100 0 R /Parent 1 0 R /Type /Page /Resources 101 0 R /MediaBox [ 0 0 612 792 ] >> endobj 12 0 obj << /Contents 104 0 R /Parent 1 0 R /Type /Page /Resources 105 0 R /MediaBox [ 0 0 612 792 ] >> endobj 13 0 obj << /Length 3020 /Filter /FlateDecode >> stream xڵYK?En? b+?ǖk??`DṙV_~٥rHNhv 4U??%?7Z@?"*??o?YGe???L?2:M}MeyeeJWCh[7nL׵{Rfg?}Á?w LTVGl?VL?2*M˼ucr ?ѰC?nn[Gi=WA?`??*w Kaddh\Dy̢LXk{?ܭNÏ?lΨ̑-ؔS?NKߚNC~8?j??,?]?vV^QQ?9lywvm"[3I?o?h_ؔQ?w~?t???Qf?6Q?`C?]dQ`?޶:?ճ{2*JX߀?/Tفa*߸??J֖nbÊo*tM7?o*V=J??&ϲA? ٶ