[B! æ©Ÿæ¢°å¦ç¿’][DeepLearning] turuhashiã®ãƒ

turuhashi id:turuhashi

æ©Ÿæ¢°å¦ç¿’ã¨DeepLearningã«é–¢ã™ã‚‹turuhashiã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (5)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®å‹•ç‰©åœ’ : ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ãƒ»ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã®ãƒãƒ¼ãƒˆã‚·ãƒ¼ãƒˆ(å‰ç·¨) | POSTD
æ–°ã—ã„ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ãŒãã®æ™‚ã€…ã§èª•ç”Ÿã™ã‚‹ãŸã‚ã€ãã‚Œã‚‰å…¨éƒ¨ã‚’æŠŠæ¡ã™ã‚‹ã“ã¨ã¯å›°é›£ã§ã™ã€‚å…¨ã¦ã®ç•¥èªžã‚’è¦šãˆã‚ˆã†ã¨ã™ã‚‹ã¨ã€æœ€åˆã¯ãã®æ•°ã®å¤šã•ã«åœ§å€’ã•ã‚Œã¦ã—ã¾ã†ã§ã—ã‚‡ã†ï¼ˆDCIGNã‚„BiLSTMã€DCGANã‚’çŸ¥ã£ã¦ã„ã‚‹äººã¯ã„ã¾ã™ã‹ï¼Ÿï¼‰ã€‚ ãã‚“ãªã‚ã‘ã§ã€ã“ã‚Œã‚‰ã®ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã®å¤šãã‚’ç››ã‚Šè¾¼ã‚“ã ãƒãƒ¼ãƒˆã‚·ãƒ¼ãƒˆã‚’ä½œã‚‹ã“ã¨ã«ã—ã¾ã—ãŸã€‚ãã®ã»ã¨ã‚“ã©ã¯ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã§ã™ã€‚ã—ã‹ã—ã€ä¸ã«ã¯å…¨ãç•°ãªã‚‹ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã‚‚æ½œã‚“ã§ã„ã¾ã™ã€‚ã©ã‚Œã‚‚ç‹¬ç‰¹ã§ç›®æ–°ã—ã„ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã°ã‹ã‚Šã§ã™ãŒã€ãƒŽãƒ¼ãƒ‰ã®æ§‹é€ ã‚’æãã“ã¨ã§åŸºæœ¬çš„ãªé–¢ä¿‚ãŒåˆ†ã‹ã‚Šã‚„ã™ããªã£ã¦ãã¾ã™ã€‚ ã“ã‚Œã‚‰ã‚’ãƒŽãƒ¼ãƒ‰ãƒžãƒƒãƒ—ã¨ã—ã¦æãã“ã¨ã®å•é¡Œç‚¹ã¯ã€ã“ã‚Œã‚‰ãŒã©ã®ã‚ˆã†ã«ä½¿ã‚ã‚Œã‚‹ã‹ã‚’æ˜Žç¢ºã«ç¤ºã—ã¦ã„ãªã„ã¨ã„ã†ç‚¹ã§ã™ã€‚ä¾‹ãˆã°ã€å¤‰åˆ†ã‚ªãƒ¼ãƒˆã‚¨ãƒ³ã‚³ãƒ¼ãƒ€ï¼ˆVAEï¼‰ã¯ã‚ªãƒ¼ãƒˆã‚¨ãƒ³ã‚³ãƒ¼ãƒ€ï¼ˆAEï¼‰ã¨åŒã˜ã‚ˆã†ã«è¦‹ãˆã¾ã™ãŒã€å®Ÿéš›ã¯è¨“ç·´éŽç¨‹ãŒå…¨ãç•°ãªã‚Šã¾ã™ã—ã€è¨“ç·´ã—ãŸãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®
turuhashi 2017/05/07
deeplearning

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
StackGANã«ã‚ˆã‚‹ãƒ•ã‚©ãƒ³ãƒˆã®éŒ¬é‡‘è¡“ - ã«ã»ã‚“ã”ã®ã‚Œã‚“ã—ã‚…ã†
StackGANã«ã‚ˆã‚‹ãƒ•ã‚©ãƒ³ãƒˆã®éŒ¬é‡‘è¡“ å›³1. ä¸å¤®ãŒéŒ¬é‡‘ã—ãŸãƒ•ã‚©ãƒ³ãƒˆ è¿‘æ³ å›³2. çœŸç†ã®ä¸€æ’ƒã‚’æ”¾ã¨ã†ã¨ã™ã‚‹ã‚«ãƒªã‚ªã‚¹ãƒˆãƒã•ã‚“ï¼ˆå…¬å¼çµµã‚ˆã‚Šï¼‰ã€€ã‚°ãƒ©ãƒ³ãƒ–ãƒ«ãƒ¼ãƒ•ã‚¡ãƒ³ã‚¿ã‚¸ãƒ¼ã¨ã„ã†ã‚¹ãƒžãƒ›ã®ã‚²ãƒ¼ãƒ ã§ã‚«ãƒªã‚ªã‚¹ãƒˆãƒã¨ã„ã†ã‚ãƒ£ãƒ©ã‚¯ã‚¿ãƒ¼ãŒã„ã¾ã™ã€‚ã‚‚ã¨ã¯å¤©æ‰éŒ¬é‡‘è¡“å¸«ã§ã€è‡ªå·±ã®æ€§åˆ¥ã‚’éŒ¬é‡‘è¡“ã§ç”·ã‹ã‚‰å¥³ã«å¤‰ãˆãŸå¤©æ‰ã§ã™ã€‚ãã‚“ãªå½¼ï¼ˆå½¼å¥³ï¼Ÿï¼‰ãŒçœŸç†ã®ä¸€æ’ƒã ãƒ¼ï¼ã£ã¦ã„ã†ã‚“ã§ã™ã‚ˆã€ã‹ã£ã“ã‚ˆã„ã§ã™ãã€‚ ã‚ã¾ã‚Šã«ã‚‚ä»Šã¾ã§ã®ä¸å¯èƒ½ã‚’æ‰“é–‹ã—ç¶šã‘ã‚‹Deep Learningã€‚è‡ªåˆ†ã§è¨€ã†ã¨ç”šã ç§‘å¦ã¨ã—ã¦ã®æ©Ÿæ¢°å¦ç¿’ãŒé›¶è½ã—ãã†ã§ã™ãŒã€ã¾ã Deep Learningã¯ä½“ç³»åŒ–ã•ã‚ŒãŸçŸ¥æµã®é›†åˆä½“ã¨ã—ã¦ã®æ£ã—ã„ç§‘å¦ã®æ®µéšŽã«ã¯ãŸã©ã‚Šç€ã„ã¦ã„ãªã„ã‚ˆã†ã«æ€ã„ã¾ã™ã€‚ã©ã¡ã‚‰ã‹ã¨è¨€ã†ã¨éŒ¬é‡‘è¡“ã«è¿‘ã„æŠ€ã‹ã‚‚ã€‚ Deep Learningã¯ã€ã„ã¤ã‹ã¾ã è¦‹ã¬çœŸç†ã¸ã¨äººé¡žã‚’å°Žã„ã¦ãã‚Œã‚‹ã‚“ã§ã—ã‚‡ã†ã‹ã€‚å„äººã€æœŸå¾…ã—ã¦ãŠã‚Šã¾ã™ã€‚ ãƒ¢ãƒãƒ™ãƒ¼ã‚·ãƒ§ãƒ³ æ—¥æœ¬èªžã®ãƒ•ã‚©ãƒ³ãƒˆã‚’ä½œæˆã™ã‚‹
turuhashi 2017/02/07
ã‚ã¨ã§èªã‚€

æ©Ÿæ¢°å¦ç¿’

deeplearning
ãƒªãƒ³ã‚¯
ã“ã“ã‹ã‚‰ã¯ã˜ã‚ã‚‹ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®ä¸»æµæ‰‹æ³•ã¾ã¨ã‚ã€ä¿å˜ç‰ˆã€‘ - HELLO CYBERNETICS
s0sem0y.hatena blog.com ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã¯2006å¹´ã«ãƒˆãƒãƒ³ãƒˆå¤§å¦ã®Hintonæ°ãŒè€ƒæ¡ˆã—ãŸDeep Belief Networkã‹ã‚‰å§‹ã¾ã‚Šã¾ã—ãŸã€‚ã“ã®ã¨ãè€ƒæ¡ˆã•ã‚ŒãŸã®ãŒäº‹å‰å¦ç¿’ã¨ã„ã†æ‰‹æ³•ã§ã€ä¸€å±¤ãšã¤å±¤ã‚’å¦ç¿’ã•ã›ã¦ã‹ã‚‰ç¹‹ã’ã¦ã„ãã“ã¨ã§ã€ä¸€æ°—ã«æ·±ã„æ§‹é€ ã‚’å¦ç¿’ã•ã›ã‚‹éš›ã«ç”Ÿã˜ã¦ã„ãŸå‹¾é…æ¶ˆå¤±å•é¡Œã‚’å›žé¿ã™ã‚‹ã“ã¨ã«æˆåŠŸã—ã¾ã—ãŸã€‚ ãã®å¾Œæ·±ã„æ§‹é€ ã‚’ä½¿ã£ãŸæ©Ÿæ¢°å¦ç¿’ãŒå…¨ã¦ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã¨å‘¼ã°ã‚Œã‚‹ã‚ˆã†ã«ãªã‚Šã€ä»Šã‚„ãã®ç¨®é¡žã¯éžå¸¸ã«å¤šå²ã«æ¸¡ã‚Šã¾ã™ã€‚ä»Šå›žã¯ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã£ã¦è¨€ã£ã¦ã‚‚è‰²ã€…å‡ºã¦ãã‚‹ã‘ã©ã€ã©ã‚ŒãŒã©ã‚Œã ã‹åˆ†ã‹ã‚‰ãªã„ã¨ã„ã†äººã®ãŸã‚ã«ï¼ˆãã—ã¦è‡ªåˆ†ã®ä¸ã§æ•´ç†ã™ã‚‹ãŸã‚ã«ï¼‰ã€ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°æ‰‹æ³•ã‚’ã¾ã¨ã‚ã¦ã¿ãŸã„ã¨æ€ã„ã¾ã™ã€‚ ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®ç¨®é¡žã®ã¾ã¨ã‚ Deep Belief Network(DBN) æœ€åˆã«ç™»å ´ã—ãŸãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®æ‰‹æ³• ç‰¹å¾´æŠ½å‡ºã‚’äººæ‰‹ã§ã‚„ã‚‰ãªãã¦ã‚‚è‰¯ããªã£ãŸ
turuhashi 2016/12/07
deeplearning

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
ã€Œã‚¼ãƒã‹ã‚‰ä½œã‚‹Deep Learningã€æ›¸è©•ã€‚è³¼å…¥ã‚’è¿·ã£ã¦ã„ã‚‹æ–¹ã¸ - HELLO CYBERNETICS
çµè«– ç†ç”± å®Ÿè£…ãŒãƒ©ã‚¤ãƒ–ãƒ©ãƒªã«ä¾å˜ã—ã¦ã„ãªã„ ã—ã£ã‹ã‚Šã¨è§£èª¬ãŒè¡Œã‚ã‚Œã¦ã„ã‚‹ ãƒªãƒ•ã‚¡ãƒ¬ãƒ³ã‚¹ãŒå……å®Ÿã—ã¦ã„ã‚‹ è¨ˆç®—ã‚°ãƒ©ãƒ•ã«ã‚ˆã£ã¦ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®è¨ˆç®—ã‚’æ‰±ã† ä¸èº« èª°ã«ãŠã™ã™ã‚ã§ãã‚‹ã‹ èª°ã«ãŠã™ã™ã‚ã§ããªã„ã‹ ç§ãŒæœ¬ã‚’è³¼å…¥ã—ã¦ çµè«– éžå¸¸ã«è‰¯ã„æ›¸ç±ã§ã™ã€‚è¿·ã£ã¦ã„ã‚‹ãªã‚‰ã°è³¼å…¥ã—ã¦ã‚‚ã„ã„ã§ã—ã‚‡ã†ã€‚ ã‚¼ãƒã‹ã‚‰ä½œã‚‹Deep Learning â€•Pythonã§å¦ã¶ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®ç†è«–ã¨å®Ÿè£… ä½œè€…: æ–Žè—¤åº·æ¯… å‡ºç‰ˆç¤¾/ãƒ¡ãƒ¼ã‚«ãƒ¼: ã‚ªãƒ©ã‚¤ãƒªãƒ¼ã‚¸ãƒ£ãƒ‘ãƒ³ ç™ºå£²æ—¥: 2016/09/24 ãƒ¡ãƒ‡ã‚£ã‚¢: å˜è¡Œæœ¬ï¼ˆã‚½ãƒ•ãƒˆã‚«ãƒãƒ¼ï¼‰ ã“ã®å•†å“ã‚’å«ã‚€ãƒ–ãƒã‚° (5ä»¶) ã‚’è¦‹ã‚‹ ç†ç”± ä»¥ä¸‹è¨˜äº‹ã§ç†ç”±ã‚’è¿°ã¹ã¦ã„ãã¾ã™ã€‚ å®Ÿè£…ãŒãƒ©ã‚¤ãƒ–ãƒ©ãƒªã«ä¾å˜ã—ã¦ã„ãªã„ ã‚‚ã—ã‚‚ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã«ä¾å˜ã—ãŸã‚‚ã®ã§ã‚ã‚‹å ´åˆã€ãƒ©ã‚¤ãƒ–ãƒ©ãƒªãŒæ›´æ–°ã•ã‚Œã‚‹ã¨ã€æœ¬ã®å†…å®¹ãŒå¤ããªã£ã¦ä¾¡å€¤ã®ãªã„ã‚‚ã®ã«ãªã‚‹ã“ã¨ã‚‚ã‚ã‚Šãˆã¾ã™ã€‚ã—ã‹ã—ã“ã®æœ¬ã¯ã€ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’ç†è§£ã™ã‚‹ã“ã¨ã«ç„¦ç‚¹ã‚’
turuhashi 2016/12/06
deeplearning

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
DeepLearningã§ä¸Šå¸ã‚’èªè˜ã—ã¦ç”»é¢ã‚’éš ã™ - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? èƒŒæ™¯ ä»•äº‹ã‚’ã—ã¦ã„ã‚‹ã¨ãã€æ¥å‹™ã«é–¢ä¿‚ãªã„æƒ…å ±ã‚’é–²è¦§ã—ã¦ã„ã‚‹ã“ã¨ã£ã¦èª°ã§ã‚‚ã‚ã‚Šã¾ã™ã‚ˆãï¼Ÿ ãã‚“ãªã¨ãã«èƒŒå¾Œã«ãƒœã‚¹ãŒå¿ã³å¯„ã£ã¦ã„ã‚‹ã¨æ°—ã¾ãšã„æ€ã„ã‚’ã—ã¾ã™ã€‚ã‚‚ã¡ã‚ã‚“æ€¥ã„ã§ç”»é¢ã‚’åˆ‡ã‚Šæ›¿ãˆã‚Œã°è‰¯ã„ã®ã§ã™ãŒã€ãã†ã„ã†å‹•ä½œã¯é€†ã«æ€ªã—ã¾ã‚Œã‚‹ã“ã¨ã«ãªã‚Šã¾ã™ã—ã€é›†ä¸ã—ã¦ã„ã‚‹ã¨ãã¯æ°—ãŒã¤ã‹ãªã„ã“ã¨ã‚‚ã‚ã‚Šã¾ã™ã€‚ãã“ã§æ€ªã—ã¾ã‚Œãšã«ç”»é¢ã‚’åˆ‡ã‚Šæ›¿ãˆã‚‹ãŸã‚ã«ã€ãƒœã‚¹ãŒè¿‘ã¥ã„ã¦ããŸã“ã¨ã‚’è‡ªå‹•çš„ã«èªè˜ã—ã¦ç”»é¢ã‚’éš ã™ã‚·ã‚¹ãƒ†ãƒ ã‚’ä½œã£ã¦ã¿ã¾ã—ãŸã€‚ å…·ä½“çš„ã«ã¯Kerasã‚’ç”¨ã„ã¦ãƒœã‚¹ã®é¡”ã‚’æ©Ÿæ¢°å¦ç¿’ã—ã€ã‚«ãƒ¡ãƒ©ã‚’ç”¨ã„ã¦è¿‘ã¥ã„ã¦ããŸã“ã¨ã‚’èªè˜ã—ã¦ç”»é¢ã‚’åˆ‡ã‚Šæ›¿ãˆã¦ã„ã¾ã™ã€‚ ãƒŸãƒƒã‚·ãƒ§ãƒ³ ãƒŸãƒƒã‚·
turuhashi 2016/09/19
æ©Ÿæ¢°å¦ç¿’

DeepLearning
ãƒªãƒ³ã‚¯
1