Deconvolutionã®ä»£ã‚ã‚Šã«Pixel Shufflerを使ã†æ¦‚è¦ Chainerã§Pixel Shufflerを実装ã—㟠ã¯ã˜ã‚㫠昨年ã«Twitter社ãŒç™ºè¡¨ã—ãŸè¶…解åƒã®è«–文をèªã‚“ã§ã„ãŸã¨ã“ã‚ã€ç”»åƒã®æ‹¡å¤§ã«ã‚ˆã用ã„られるDeconvolutionを使ã‚ãšã€Pixel Shufflerã¨å‘¼ã°ã‚Œã‚‹ä»•çµ„ã¿ã‚’利用ã—ã¦è¶…解åƒã‚’è¡Œã£ã¦ãŠã‚Šã€æ°—ã«ãªã£ãŸã®ã§èª¿ã¹ã¾ã—ãŸã€‚ Pixel Shufflerã¯Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Networkã§æ案ã•ã‚ŒãŸã‚‚ã®ã§ï¼ˆæ£å¼ã«ã¯Sub-Pixel Convolution）ã€å…¥åŠ›ç‰¹å¾´ãƒžãƒƒãƒ—ã®å„ピクセルを並ã¹æ›¿ãˆã¦é«˜è§£åƒåº¦ãªç‰¹å¾´ãƒžãƒƒãƒ—を出力ã—ã¾ã™ã€‚ 今回ã¯DCGANã®Generatorã‚’Deconvolutionã¨Pixel Shufflerã®ä¸¡æ–¹ã§æ§‹æˆã—
{{#tags}}- {{label}}
{{/tags}}