[B! æ©Ÿæ¢°å¦ç¿’][dl] nakamodsã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

nakamods id:nakamods

æ©Ÿæ¢°å¦ç¿’ã¨dlã«é–¢ã™ã‚‹nakamodsã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (15)

${{author_name}}$

{{{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}}

ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿è‡ªå‹•æœ€é©åŒ–ãƒ„ãƒ¼ãƒ«ã€ŒOptunaã€å…¬é–‹ - Preferred Networks Research & Development
ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿è‡ªå‹•æœ€é©åŒ–ãƒ•ãƒ¬ãƒ¼ãƒ ãƒ¯ãƒ¼ã‚¯ã€ŒOptunaã€ã®ãƒ™ãƒ¼ã‚¿ç‰ˆã‚’ OSS ã¨ã—ã¦å…¬é–‹ã—ã¾ã—ãŸã€‚ã“ã®è¨˜äº‹ã§ã¯ã€Optuna ã®é–‹ç™ºã«è‡³ã£ãŸå‹•æ©Ÿã‚„ç‰¹å¾´ã‚’ç´¹ä»‹ã—ã¾ã™ã€‚ å…¬å¼ãƒšãƒ¼ã‚¸ å…¬å¼ãƒ‰ã‚ãƒ¥ãƒ¡ãƒ³ãƒˆ ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ« GitHub ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã¨ã¯ï¼Ÿ ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã¨ã¯ã€æ©Ÿæ¢°å¦ç¿’ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®æŒ™å‹•ã‚’åˆ¶å¾¡ã™ã‚‹ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã®ã“ã¨ã§ã™ã€‚ç‰¹ã«æ·±å±¤å¦ç¿’ã§ã¯å‹¾é…æ³•ã«ã‚ˆã£ã¦æœ€é©åŒ–ã§ããªã„ãƒ»ã—ãªã„ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã«ç›¸å½“ã—ã¾ã™ã€‚ä¾‹ãˆã°ã€å¦ç¿’çŽ‡ã‚„ãƒãƒƒãƒã‚µã‚¤ã‚ºã€å¦ç¿’ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³æ•°ã¨ã„ã£ãŸã‚ˆã†ãªã‚‚ã®ãŒãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã¨ãªã‚Šã¾ã™ã€‚ã¾ãŸã€ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®å±¤æ•°ã‚„ãƒãƒ£ãƒ³ãƒãƒ«æ•°ã¨ã„ã£ãŸã‚ˆã†ãªã‚‚ã®ã‚‚ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã§ã™ã€‚æ›´ã«ã€ãã®ã‚ˆã†ãªæ•°å€¤ã ã‘ã§ãªãã€å¦ç¿’ã« Momentum SGD ã‚’ç”¨ã„ã‚‹ã‹ãã‚Œã¨ã‚‚ Adam ã‚’ç”¨ã„ã‚‹ã‹ã€ã¨ã„ã£ãŸã‚ˆã†ãªé¸æŠžã‚‚ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã¨è¨€ãˆã¾ã™ã€‚ ãƒã‚¤ãƒ‘ãƒ¼ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã®èª¿æ•´ã¯æ©Ÿæ¢°å¦ç¿’ã‚¢
nakamods 2018/12/03
chainer

æ©Ÿæ¢°å¦ç¿’

dl

python
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®åˆ¤æ–æ ¹æ‹ ã‚’ç†è§£ã™ã‚‹æ‰‹æ³• - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã¯ç‰¹å®šåˆ†é‡Žã§éžå¸¸ã«é«˜ã„ç²¾åº¦ãŒå‡ºã›ã‚‹ã“ã¨ã‚‚ã‚ã‚Šã€ãã®å¿œç”¨ç¯„å›²ã¯ã©ã‚“ã©ã‚“åºƒãŒã£ã¦ã„ã¾ã™ã€‚ ã—ã‹ã—ã€ãã‚“ãªãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã«ã‚‚å¼±ç‚¹ã¯ã‚ã‚Šã¾ã™ã€‚ãã®ä¸ã§ã‚‚å¤§ãã„å•é¡Œç‚¹ãŒã€ã€Œä½•ã‚’æ ¹æ‹ ã«åˆ¤æ–ã—ã¦ã„ã‚‹ã‹ã‚ˆãã‚ã‹ã‚‰ãªã„ã€ã¨ã„ã†ã“ã¨ã§ã™ã€‚ ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã¯ã€å¦ç¿’ã®éŽç¨‹ã§ãƒ‡ãƒ¼ã‚¿å†…ã®ç‰¹å¾´ãã‚Œè‡ªä½“ã‚’å¦ç¿’ã™ã‚‹ã®ãŒå¾—æ„ã¨ã„ã†ç‰¹æ€§ãŒã‚ã‚Šã¾ã™ã€‚ã“ã‚Œã«ã‚ˆã‚Šã€ŒäººãŒç‰¹å¾´ã‚’æŠ½å‡ºã™ã‚‹å¿…è¦ãŒãªã„ã€ã¨è¨€ã‚ã‚ŒãŸã‚Šã‚‚ã—ã¾ã™ãŒã€é€†ã«è¨€ãˆã°ã©ã‚“ãªç‰¹å¾´ã‚’æŠ½å‡ºã™ã‚‹ã‹ã¯ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ä»»ã›ã¨ã„ã†ã“ã¨ã§ã™ã€‚æŠ½å‡ºã•ã‚ŒãŸç‰¹å¾´ã¯ãã®åã®é€šã‚Šãƒ‡ã‚£ãƒ¼ãƒ—ãªãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®ä¸ã®é‡ã¿ã«æ½œåœ¨ã—ã¦ãŠ
nakamods 2017/09/06
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
Deep Learning
This domain may be for sale!
nakamods 2017/09/06
AI

æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
æ©Ÿæ¢°å¦ç¿’ã‚’ã‚„ã‚‹ã¨ã€AIã‚ˆã‚Šäººé–“ã®æ–¹ãŒè©³ã—ããªã‚‹ç¾è±¡ - bohemiaæ—¥è¨˜
ã“ã‚“ã«ã¡ã¯ã€‚ã¼ã¸ã¿ã‚ã§ã™ã€‚ æ©Ÿæ¢°å¦ç¿’ãŒã™ã£ã‹ã‚Šãƒ–ãƒ¼ãƒ ã«ãªã£ã¦ã€ä»•äº‹ã‚„è¶£å‘³ã§ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’ä½¿ã£ãŸã“ã¨ãŒã‚ã‚‹äººãŒå¢—ãˆã¦ã„ã‚‹ã¨æ€ã„ã¾ã™ã€‚ ç‰¹ã«ç”»åƒåˆ†é‡Žã§ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã¯æˆæžœã‚’ä¸Šã’ã¦ã„ã‚‹ã®ã§ã€ç‰¹å®šã®ã‚‚ã®ã‚’åˆ¤åˆ¥ãƒ»è˜åˆ¥ã™ã‚‹ã¨ã„ã£ãŸäº‹ä¾‹ãŒå¤šã„ã‹ã¨æ€ã„ã¾ã™ã€‚ ãã†ã—ãŸç”»åƒç³»ã®ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°çµŒé¨“è€…ãªã‚‰çµŒé¨“ã™ã‚‹ç¾è±¡ãŒã‚ã‚Šã¾ã™ã€‚ ãã‚Œã¯ã€ ã€Œãã®åˆ†é‡Žã«ã¤ã„ã¦ã€ä½œã£ãŸAIã‚ˆã‚Šã‚‚è‡ªåˆ†ãŒè©³ã—ããªã‚‹ã€ ã¨ã„ã†ç¾è±¡ã§ã™ã€‚ ãŠãæ¾ã•ã‚“è¦‹åˆ†ã‘ã®ç¬¬ä¸€äººè€…ã«ãªã£ã¦ã—ã¾ã£ãŸ ä»¥å‰å–ã‚Šçµ„ã‚“ã ãŠãæ¾ã•ã‚“ã®6ã¤åã‚’ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§è¦‹åˆ†ã‘ã‚‹ãƒã‚¿ã§ã¯ã€å¦ç¿’ç”¨ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã®ä½œæˆã®ãŸã‚ã«ã€è‡ªåŠ›ã§5000æžšä»¥ä¸Šã®ãŠãæ¾ã•ã‚“ãŸã¡ã‚’è¦‹åˆ†ã‘ã‚‹ä½œæ¥ã‚’è¡Œã„ã¾ã—ãŸã€‚ãã®çµæžœã€ã‚¢ãƒ‹ãƒ¡ã‚’è¦‹ã¦ã‚‚ã€ã“ã‚Œã¯ä½•æ¾ã ã€ã¨åˆ¤æ–ã§ãã‚‹ã‚ˆã†ã«ãªã£ã¦ã—ã¾ã„ã¾ã—ãŸã€‚å½“æ™‚ã¯ãŠãæ¾ã•ã‚“ã‚’è¦‹åˆ†ã‘ã‚‹èƒ½åŠ›ã¯èª°ã«ã‚‚è² ã‘ãªã„è‡ªä¿¡ãŒã‚ã‚Šã¾ã—ãŸã€‚ ãªãœã‹å¦ä¼šã§ãŠãæ¾ã•ã‚“ã®è©±ã‚’ã™ã‚‹æ©Ÿ
nakamods 2017/03/23
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
å¹³æˆç”Ÿã¾ã‚ŒãŒä½œã‚‹ã€Œæ†§ã‚Œã®80å¹´ä»£ã€ ç„¡åã®ãƒ¦ãƒ‹ãƒƒãƒˆãŒã‚¢ãƒ¡ãƒªã‚«ã®éŸ³æ¥½ãƒ•ã‚§ã‚¹ã«å‘¼ã°ã‚Œã‚‹ã¾ã§
Search, watch, and cook every single Tasty recipe and video ever - all in one place! News, Politics, Culture, Life, Entertainment, and more. Stories that matter to you.
nakamods 2017/03/12
dl

æ©Ÿæ¢°å¦ç¿’

éŸ³æ¥½

æ˜ åƒ
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§ã‚¢ã‚¹ã‚ãƒ¼ã‚¢ãƒ¼ãƒˆã‚’ä½œã‚‹ - Qiita
ã¯ã˜ã‚ã¾ã—ã¦ã€‚ æœ¬æ¥ã¯ã‚¢ã‚¹ã‚ãƒ¼ã‚¢ãƒ¼ãƒˆ (ä»¥ä¸‹AA) è·äººã®OsciiArtã¨ã„ã„ã¾ã™ (æœ¬æ¥ã§ã¯ãªã„)ã€‚ Alpha Goå¯¾ã‚¤ãƒ»ã‚»ãƒ‰ãƒ«ã®å¯¾å±€ã‚’è¦‹ã¦ã€ã€Œåƒ•ã‚‚ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§ç¥žAAè·äººã‚’å€’ã—ãŸã„ï¼ã€ã¨æ€ã„ã€pythonã‚’ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã—ã¦ã¡ã‚‡ã†ã©ä¸€å¹´ã®æˆæžœã‚’æ›¸ã„ã¦ã„ãã¾ã™ã€‚ ã‚³ãƒ¼ãƒ‰ã¯ã“ã¡ã‚‰ã«ã‚¢ãƒƒãƒ—ã—ã¦ã‚ã‚Šã¾ã™ã€‚ https://github.com/OsciiArt/DeepAA ã“ã“ã§æ‰±ã†ã‚¢ã‚¹ã‚ãƒ¼ã‚¢ãƒ¼ãƒˆã¨ã¯ ã“ã“ã§æ‰±ã†AAã¨ã¯ã€ ã“ã†ã„ã†ã®â€¦â€¦â†“ ã§ã¯ãªãã€ã“ã†ã„ã†ã®â€¦â€¦â†“ ã§ã‚‚ãªãã€ã“ã†ã„ã†ã®â€¦â€¦â†“ ã¨ã‚‚ã¡ã‚‡ã£ã¨é•ã£ã¦ã€ã“ã†ã„ã†ã®â€¦â€¦â†“ ã§ã¯ã‚‚ã¡ã‚ã‚“ãªãã€ã“ã†ã„ã†ã®ã§ã™ã€‚â†“ ã“ã®ã‚ˆã†ãªã€ç·šç”»ã‚’æ–‡å—ã‚’ä½œã£ã¦å†ç¾ã—ãŸã€Œãƒˆãƒ¬ãƒ¼ã‚¹AAã€ã¨å‘¼ã°ã‚Œã‚‹ã‚¿ã‚¤ãƒ—ã®AAã‚’ã“ã“ã§ã¯æ‰±ã„ã¾ã™ã€‚ è©³ç´°ã¯wikipediaã®ã€Œã‚¢ã‚¹ã‚ãƒ¼ã‚¢ãƒ¼ãƒˆã€ã®ãƒšãƒ¼ã‚¸ã®ã€Œãƒ—ãƒãƒãƒ¼ã‚·ãƒ§ãƒŠãƒ«ãƒ•ã‚©ãƒ³ãƒˆã€ã®é …ã‚’å‚ç…§ã—ã¦ãã ã•ã„ã€‚ wikipe
nakamods 2017/03/12
dl

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
20170127-learning-tensorflow-3hours
ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’ç”¨ã„ãŸæ©Ÿæ¢°å¦ç¿’ã®ã€Œãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°(æ·±å±¤å¦ç¿’)ã€ã¯äººå·¥çŸ¥èƒ½(AI)é–‹ç™ºã«æ¬ ã‹ã›ãªã„æŠ€è¡“ã§ã‚ã‚Šã€AIä»¥å¤–ã«ã‚‚Googleãƒ•ã‚©ãƒˆã®ç”»åƒã®è‡ªå‹•ã‚¿ã‚°ä»˜ã‘æ©Ÿèƒ½ã‚„Amazonã®ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰æ©Ÿèƒ½ãªã©ã€ã™ã§ã«å®Ÿç”¨åŒ–ã•ã‚Œã¦ã„ã‚‹æŠ€è¡“ã«æ´»ç”¨ã•ã‚Œã¦ã„ã¾ã™ã€‚ãã‚“ãªãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’é–‹ç™ºè€…ãŒå¦ç¿’ã™ã‚‹ãŸã‚ã®ã¨ã£ã¦ãŠãã®æ–¹æ³•ã‚’Googleã®ã‚¯ãƒ©ã‚¦ãƒ‰é–‹ç™ºè€…ãŒãƒ–ãƒã‚°ã§ç´¹ä»‹ã—ã¦ã„ã¾ã™ã€‚ Learn TensorFlow and deep learning, without a Ph.D. | Google Cloud Big Data and Machine Learning Blog Â |Â Google Cloud Platform https://cloud.google.com/blog/big-data/2017/01/learn-tensorflow-and-deep-learning-w
nakamods 2017/01/27
dl

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
ã€Day-17ã€‘DeepLearningç³»ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã€ã€ŽKerasã€ã®ä½¿ã„æ–¹ã¾ã¨ã‚(2.xå¯¾å¿œç‰ˆ) - ãƒ—ãƒã‚¯ãƒ©ã‚·ã‚¹ãƒˆ
ã€æœ€çµ‚æ›´æ–° : 2017.12.17ã€‘ â€»ä»¥å‰æ›¸ã„ãŸè¨˜äº‹ãŒObsoleteã«ãªã£ãŸãŸã‚ã€2.xã§ãã¡ã‚“ã¨å‹•ãã‚ˆã†ã«æ›¸ãç›´ã—ã¾ã—ãŸã€‚ ãƒ‡ãƒ¼ã‚¿åˆ†æžã‚¬ãƒå‹‰å¼·ã‚¢ãƒ‰ãƒ™ãƒ³ãƒˆã‚«ãƒ¬ãƒ³ãƒ€ãƒ¼ 17æ—¥ç›®ã€‚ 16æ—¥ç›®ã«ã€1ã‹ã‚‰ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆã‚’æ›¸ãã¾ã—ãŸã€‚ ãã‚Œã¯ãã‚Œã§ãƒ‡ãƒ¼ã‚¿ã®æµã‚Œã ã¨ã‹ã€æ´»æ€§åŒ–é–¢æ•°ã®åƒãã ã¨ã‹å¾—ã‚‹ã‚‚ã®ã¯å¤šã‹ã£ãŸã®ã§ã™ãŒã€Kerasã¨è¨€ã†ã‚‚ã®ã‚’ä½¿ã£ã¦ã¿ã¦ã€ä½•ã¦ç´ æ™´ã‚‰ã—ã„ã‚“ã ã¨æ„Ÿå‹•ã—ã¦ã—ã¾ã„ã¾ã—ãŸ ä»Šã¾ã§è‹¦åŠ´ã—ã¦æ•°åè¡Œæ›¸ã„ã¦ã„ãŸã‚‚ã®ãŒã€ã‚ãšã‹3è¡Œã§æ›¸ã‘ã‚‹ï¼ æ£ç›´ã€ã‚¹ã‚¯ãƒ©ãƒƒãƒã§æ›¸ãæ„å‘³ã£ã¦ã€ç†è§£ã«ã¯ã„ã„ã‘ã©ç ”ç©¶ã‚„åˆ†æžã«ã¯å¿…è¦ã‚ã‚“ã¾ãªã„ã‚“ã§ã™ã‚ˆãã€‚è»Šè¼ªã®å†ç™ºæ˜Žã«ãªã‚‹ã—ã€‚ ã¨è¨€ã†ã‚ã‘ã§ã€ä½¿ãˆã‚‹ã‚‚ã®ã¯ã©ã‚“ã©ã‚“ä½¿ã£ã¦ã„ã“ã†ã‚¹ã‚¿ãƒ³ã‚¹ã§ã€ä»Šæ—¥ã¯Kerasã®ç´¹ä»‹ã§ã™ï¼ Tutorial+æ°—ã«ãªã£ãŸå¼•æ•°ã‚’æŽ˜ã‚Šä¸‹ã’ã¦è£œè¶³ã®ã‚ˆã†ãªæ„Ÿã˜ã§æ›¸ã„ã¦ã„ã¾ã™ã€‚ ã¡ãªã¿ã«ã€å„éƒ¨ã®ã‚³ãƒ¼ãƒ‰ä»¥ä¸‹ã‚’ã¤ãªãŽåˆã‚ã›ã‚‹ã¨mnistã®åˆ†é¡žå™¨ãŒå‹•ãã‚ˆ
nakamods 2017/01/08
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
åˆå¿ƒè€…ãŒchainerã§ç·šç”»ç€è‰²ã—ã¦ã¿ãŸã€‚ã‚ã‚Šã¨ã§ããŸã€‚
ï¾ƒï¾žï½°ï¾Œï¾Ÿï¾—ï½°ï¾†ï¾ï½¸ï¾žã¯ã‚³ãƒ¢ãƒ‡ã‚£ãƒ†ã‚£åŒ–ã—ã¦ã„ã¦ãƒãƒ³ãƒ€ä»˜ã‘ã®æ–¹ãŒä»˜åŠ ä¾¡å€¤é«˜ã„ã¨ã„ã‚ã‚Œã‚‹ï¾‹ï¾Ÿ-FNã§ä¸»ã«å·¥ä½œæ‹…å½“ã®tai2anã§ã™ã€‚ NHKã§å…¨å›½æ”¾é€ã•ã‚ŒãŸAmazon Picking Challengeã§ã‚¬ãƒ ãƒ†ã¹ã£ãŸã¹ãŸã®ãƒãƒ³ãƒ‰ã‚„ãƒãƒœã‚³ãƒ³æ„Ÿæº€è¼‰ã®æ»‘ã‚Šå°ã¨ã‹ã‚’å·¥ä½œã—ã¦ã¾ã—ãŸã€‚ ã¨ã¯ã„ãˆã€ã‚„ã£ã±ã‚Šã¡ã‚‡ã£ã¨ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã—ã¦ã¿ãŸã„ã®ã§1,2ã‹æœˆå‰ã‹ã‚‰chainerã‚’å‹‰å¼·ã—å§‹ã‚ã¾ã—ãŸã€‚ ã›ã£ã‹ããªã®ã§ç·šç”»ã®ç€è‰²ã‚’ã—ãŸã„ãªãƒ¼ã¨æ€ã£ã¦è‰²ã€…è©¦ã—ã¦ã¿ã¾ã—ãŸã€‚ ç·šç”»ã®ç€è‰²ã¯æ•™å¸«ã‚ã‚Šå¦ç¿’ãªã®ã§ç·šç”»ã¨ç€è‰²æ¸ˆã¿ã®ç”»åƒã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆãŒ(ã§ãã‚Œã°å¤§é‡ã«)å¿…è¦ã§ã™ã€‚ ä»Šå›žã¯OpenCVã§ã‚«ãƒ©ãƒ¼ã®ç”»åƒã‹ã‚‰ç·šç”»ã‚’é©å½“ã«æŠ½å‡ºã—ã¦ã„ã¾ã™ã€‚ æŠ½å‡ºä¾‹ â†’ ã‚«ãƒ©ãƒ¼ã®ç”»åƒã‚’é›†ã‚ã¦ç·šç”»ã‚’ä½œã‚Œã°ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã®å®Œæˆã§ã™ã€‚ï¼ˆä»Šå›žã¯ï¼–ï¼ä¸‡æžšãã‚‰ã„ä½¿ã£ã¦ã„ã¾ã™ï¼‰ ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®å½¢ã§ã™ãŒã€U-netã¨ã„ã†æœ€åˆã®æ–¹ã§ã‚³ãƒ³ãƒœãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã™ã‚‹æ™‚ã®å±¤ã®å‡º
nakamods 2016/12/26
chainer

Python

æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
åºœå¤§ç”ŸãŒè¶£å‘³ã§ä¸–ç•Œä¸€ã®èªè˜ç²¾åº¦ã‚’æŒã¤ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’é–‹ç™ºã—ã¦ã—ã¾ã£ãŸè«–æ–‡ã‚’èªã‚“ã - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? Deep Learning Advent Calendar 2016ã®20æ—¥ç›®ã®è¨˜äº‹ã§ã™ã€‚ ConvNetã®æ´å²ã¨ResNetäºœç¨®ã€ãƒ˜ã‚™ã‚¹ãƒˆãƒ•ã‚šãƒ©ã‚¯ãƒ†ã‚£ã‚¹ã«é–¢é€£ã‚¹ãƒ©ã‚¤ãƒ‰ãŒã‚ã‚Šã¾ã™ï¼ˆè¿½è¨˜ï¼‰ èƒŒæ™¯ åºœå¤§ç”ŸãŒè¶£å‘³ã§ä¸–ç•Œä¸€ã®èªè˜ç²¾åº¦ã‚’æŒã¤ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’é–‹ç™ºã—ã¦ã—ã¾ã£ãŸã‚ˆã†ã§ã™ã€‚ M2ã®å¦ç”ŸãŒè¶£å‘³ã§ã‚„ã£ã¦ã„ãŸCIFAR10ã¨CIFAR100ã®èªè˜ã‚¿ã‚¹ã‚¯ã§ï¼Œç¾æ™‚ç‚¹ã§ã®ä¸–ç•Œæœ€é«˜æ€§èƒ½ã®çµæžœã‚’å‡ºã—ãŸãã†ã â€¦è¶£å‘³ã§ã£ã¦ã„ã†ã®ãŒâ€¦https://t.co/HKFLXTMbzx â€” ãƒ‹ãƒ¼ã‚·ã‚§ã‚¹ (@lachesis1120) 2016å¹´12æœˆ7æ—¥ åºœ
nakamods 2016/12/21
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§ã€Œã‚¤ãƒ³ãƒ‰äººã‚’å³ã«ã€ã‚’ç†è§£ã™ã‚‹: Generative Adversarial Network ã«ã‚ˆã‚‹ç”»åƒãƒ¢ãƒ¼ãƒ•ã‚£ãƒ³ã‚° - åŠ è—¤å¤§æ™´ã®ã‚¦ã‚§ãƒ–ã‚µã‚¤ãƒˆ
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§ã€Œã‚¤ãƒ³ãƒ‰äººã‚’å³ã«ã€ã‚’ç†è§£ã™ã‚‹: Generative Adversarial Network ã«ã‚ˆã‚‹ç”»åƒãƒ¢ãƒ¼ãƒ•ã‚£ãƒ³ã‚° ã€Œã‚¤ãƒ³ãƒ‰äººã‚’å³ã«ã€å•é¡Œ ã‚¤ãƒ³ã‚¿ãƒ¼ãƒãƒƒãƒˆã‚’é•·ãä½¿ã£ã¦ã„ã‚‹æ–¹ã¯ã€ä¼èª¬ã®èª¤æ¤ã€Œã‚¤ãƒ³ãƒ‰äººã‚’å³ã«ã€ [1] ã«ã¤ã„ã¦ã”å˜çŸ¥ãªã®ã§ã¯ãªã„ã§ã—ã‚‡ã†ã‹ã€‚ ã€ŒããŠã€œ!! ã¶ã¤ã‹ã‚‹ã€œ!! ã“ã“ã§ã‚¢ã‚¯ã‚»ãƒ«å…¨é–‹ã€ã‚¤ãƒ³ãƒ‰äººã‚’å³ã«!ã€ ã“ã®å”çªã«éŽãŽã‚‹æ„å‘³ä¸æ˜Žãªã€Œã‚¤ãƒ³ãƒ‰äººã€ã¯ã€Œãƒãƒ³ãƒ‰ãƒ«ã€ã®èª¤æ¤ã§ã‚ã‚Šã€ãã‚Œã¯ãƒ©ã‚¤ã‚¿ãƒ¼ã®æ‰‹æ›¸ãã®æ–‡å—ãŒæ±šã‹ã£ãŸãŸã‚ã«ç™ºç”Ÿã—ãŸã¨ã•ã‚Œã¦ã„ã¾ã™ã€‚ â€¦æ‰‹æ›¸ãã®æ–‡å—ãŒæ±šã‹ã£ãŸã¨ã—ã¦ã€ã©ã†ã™ã‚Œã°ã€Œãƒãƒ³ãƒ‰ãƒ«ã€ãŒã€Œã‚¤ãƒ³ãƒ‰äººã€ã«ãªã‚‹ã®ã‹ï¼Ÿã€€å¾“æ¥ã‚ˆã‚Šã€ã“ã®å•é¡Œã«ã¤ã„ã¦æ§˜ã€…ãªè€ƒå¯ŸãŒãªã•ã‚Œã¦ãã¾ã—ãŸã€‚ ã“ã“ã§ã¯ã€è¿‘å¹´ã®æŠ€è¡“ã®ç™ºå±•ã®æˆæžœã‚’å–ã‚Šå…¥ã‚Œã€ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã«æ–‡å—ã‚’æ›¸ã‹ã›ã‚‹ã“ã¨ã«ã‚ˆã£ã¦ã€Œãƒãƒ³ãƒ‰ãƒ«ã€ã‹ã‚‰ã€Œã‚¤ãƒ³ãƒ‰äººã€ã¸ã®å¤‰å®¹ã«ã¤ã„ã¦ã‚¢ãƒ—ãƒãƒ¼ãƒã—ã¦ã¿ãŸã„ã¨æ€ã„ã¾ã™ã€‚ # ã“ã‚Œã¯ De
nakamods 2016/12/07
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
ç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®ä»•çµ„ã¿ | POSTD
(ç·¨æ³¨ï¼š2016/11/17ã€è¨˜äº‹ã‚’ä¿®æ£ã„ãŸã—ã¾ã—ãŸã€‚) ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®åˆ†é‡Žã§ãƒ†ã‚¯ãƒŽãƒã‚¸ã®é€²åŒ–ãŒç¶šã„ã¦ã„ã‚‹ã¨ã„ã†ã“ã¨ãŒè©±é¡Œã«ãªã‚‹å ´åˆã€åä¸å…«ä¹ç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ãŒé–¢ä¿‚ã—ã¦ã„ã¾ã™ã€‚ç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã¯CNNï¼ˆConvolutional Neural Networkï¼‰ã¾ãŸã¯ConvNetã¨ã‚‚å‘¼ã°ã‚Œã€ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®åˆ†é‡Žã®ä¸»åŠ›ã¨ãªã£ã¦ã„ã¾ã™ã€‚CNNã¯ç”»åƒã‚’è¤‡æ•°ã®ã‚«ãƒ†ã‚´ãƒªã«åˆ†é¡žã™ã‚‹ã‚ˆã†å¦ç¿’ã—ã¦ãŠã‚Šã€ãã®åˆ†é¡žèƒ½åŠ›ã¯äººé–“ã‚’ä¸Šå›žã‚‹ã“ã¨ã‚‚ã‚ã‚Šã¾ã™ã€‚å¤§è¨€å£®èªžã®ã†ãŸã„æ–‡å¥ã‚’å®Ÿç¾ã—ã¦ã„ã‚‹æ–¹æ³•ãŒæœ¬å½“ã«ã‚ã‚‹ã¨ã™ã‚Œã°ã€ãã‚Œã¯CNNã§ã—ã‚‡ã†ã€‚ CNNã®éžå¸¸ã«å¤§ããªé•·æ‰€ã¨ã—ã¦ã€ç†è§£ã—ã‚„ã™ã„ã“ã¨ãŒæŒ™ã’ã‚‰ã‚Œã¾ã™ã€‚å°‘ãªãã¨ã‚‚å¹¾ã¤ã‹ã®åŸºæœ¬çš„ãªéƒ¨åˆ†ã«ãƒ–ãƒ¬ãƒ¼ã‚¯ãƒ€ã‚¦ãƒ³ã—ã¦å¦ã¹ã°ã€ãã‚Œã‚’å®Ÿæ„Ÿã§ãã‚‹ã§ã—ã‚‡ã†ã€‚ã¨ã„ã†ã‚ã‘ã§ã€ã“ã‚Œã‹ã‚‰ä¸€é€šã‚Šèª¬æ˜Žã—ã¾ã™ã€‚ã¾ãŸã€ç”»åƒå‡¦ç†ã«ã¤ã„ã¦ã“ã®è¨˜äº‹ã‚ˆã‚Šã‚‚è©³ç´°ã«èª¬æ˜Ž
nakamods 2016/11/18
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã«ã‚ˆã‚‹ãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã‚¢ã‚¤ãƒ†ãƒ æ¤œå‡ºã¨æ¤œç´¢ - ZOZO Technologies TECH BLOG
ãƒ‡ãƒ¼ã‚¿ã‚µã‚¤ã‚¨ãƒ³ãƒ†ã‚£ã‚¹ãƒˆã®ä¸æ‘ã§ã™ã€‚VASILYã§ã¯ãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã«ç‰¹åŒ–ã—ãŸç”»åƒè§£æžã‚¨ãƒ³ã‚¸ãƒ³ã‚’é–‹ç™ºã—ã¦ã„ã¾ã™ã€‚æœ¬è¨˜äº‹ã§ã¯ã€ã‚¹ãƒŠãƒƒãƒ—å†™çœŸã‹ã‚‰ãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã‚¢ã‚¤ãƒ†ãƒ ã‚’æ¤œå‡ºã™ã‚‹ã‚·ã‚¹ãƒ†ãƒ ã‚’ç´¹ä»‹ã—ãŸã„ã¨æ€ã„ã¾ã™ã€‚ æ¦‚è¦ ã“ã®ã‚·ã‚¹ãƒ†ãƒ ã®å…¥åŠ›ã¯ã‚¹ãƒŠãƒƒãƒ—å†™çœŸã§ã™ã€‚ã‚¹ãƒŠãƒƒãƒ—å†™çœŸãŒå…¥åŠ›ã•ã‚ŒãŸã¨ãã€ã‚·ã‚¹ãƒ†ãƒ ã¯ä»¥ä¸‹ã®ã‚¿ã‚¹ã‚¯ã‚’è§£ãã¾ã™ã€‚ å†™çœŸä¸ã‹ã‚‰ãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã‚¢ã‚¤ãƒ†ãƒ ã«è©²å½“ã™ã‚‹é ˜åŸŸã‚’æ¤œå‡ºã™ã‚‹ æ¤œå‡ºã—ãŸãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã‚¢ã‚¤ãƒ†ãƒ ã®ã‚«ãƒ†ã‚´ãƒªã‚’äºˆæ¸¬ã™ã‚‹ æ¤œå‡ºã—ãŸãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã‚¢ã‚¤ãƒ†ãƒ ã«ä¼¼ã¦ã„ã‚‹ã‚¢ã‚¤ãƒ†ãƒ ã‚’DBã‹ã‚‰æ¤œç´¢ã™ã‚‹ å„ã‚¿ã‚¹ã‚¯ã‚’è§£ãæ–¹æ³•ã¯æ§˜ã€…ã‚ã‚Šã¾ã™ãŒã€å¼Šç¤¾ã®ã‚·ã‚¹ãƒ†ãƒ ã§ã¯2ç¨®é¡žã®ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’ä½¿ã£ã¦ã“ã‚Œã‚’é”æˆã—ã¦ã„ã¾ã™ã€‚ ãƒ•ã‚¡ãƒƒã‚·ãƒ§ãƒ³ã‚¢ã‚¤ãƒ†ãƒ ã®æ¤œå‡ºã¨ã‚«ãƒ†ã‚´ãƒªäºˆæ¸¬ æ¤œå‡ºã¯ç”»åƒèªè˜ã®åŸºæœ¬çš„ãªã‚¿ã‚¹ã‚¯ã§ç››ã‚“ã«ç ”ç©¶ã•ã‚Œã¦ã„ã¦æ§˜ã€…ãªæ‰‹æ³•ãŒææ¡ˆã•ã‚Œã¦ã„ã¾ã™ãŒã€ä»Šå›žã¯Single Shot MultiBox Detector (SSD)*1 ã¨å‘¼ã°
nakamods 2016/10/25
æ©Ÿæ¢°å¦ç¿’

dl
ãƒªãƒ³ã‚¯
è·å ´ã§é–²è¦§æ³¨æ„ãªç”»åƒã‚’è‡ªå‹•ç”Ÿæˆã™ã‚‹ã‚·ã‚¹ãƒ†ãƒ ã‚’Yahooã®ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§é–‹ç™º
Yahooã¯NSFW(è·å ´é–²è¦§ä¸é©åˆ‡)ãªç”»åƒã‚’èªè˜ã™ã‚‹ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã€ŒOpen nsfwã€ã‚’ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã¨ã—ã¦å…¬é–‹ã—ã¦ã„ã‚‹ã®ã§ã™ãŒã€ã“ã‚Œã‚’ä½¿ã£ã¦ã€Œã‚³ãƒ³ã‚µãƒ¼ãƒˆã€ã€Œã‚®ãƒ£ãƒ©ãƒªãƒ¼ã€ãªã©ã®ã‚«ãƒ†ã‚´ãƒªã”ã¨ã«NSFWãªç”»åƒã‚’è‡ªå‹•ç”Ÿæˆã§ãã‚‹ã‚¸ã‚§ãƒãƒ¬ãƒ¼ã‚¿ãƒ¼ã‚’é–‹ç™ºã—ãŸçŒ›è€…ãŒç™»å ´ã—ã¦ã„ã¾ã™ã€‚ Image Synthesis from Yahoo's open_nsfw https://open_nsfw.gitlab.io/ åšå£«è«–æ–‡æå‡ºå¿—é¡˜è€…ã®Gabriel Gohæ°ã¯ã€Yahooã®ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã€ŒOpen nsfwã€ã«ã€ãƒ¯ã‚¤ã‚ªãƒŸãƒ³ã‚°å¤§å¦ã®Evolving AI LabãŒé–‹ç™ºã—ãŸç”»åƒç”Ÿæˆãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã€ŒSynthesizingã€ã‚’çµ„ã¿åˆã‚ã›ã‚‹ã“ã¨ã§ã€NSFWãªç”»åƒã‚’ç”Ÿæˆã™ã‚‹ã‚¸ã‚§ãƒãƒ¬ãƒ¼ã‚¿ãƒ¼ã‚’é–‹ç™ºã—ã¾ã—ãŸã€‚ Synthesizingã¯ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’é€šã˜ã¦ç”»åƒã‚’ç”Ÿæˆã™ã‚‹
nakamods 2016/10/22
AI

dl

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
WaveNet: A Generative Model for Raw Audio | DeepMind
Research WaveNet: A generative model for raw audio Published 8 September 2016 Authors AÃ¤ron van den Oord, Sander Dieleman This post presents WaveNet, a deep generative model of raw audio waveforms. We show that WaveNets are able to generate speech which mimics any human voice and which sounds more natural than the best existing Text-to-Speech systems, reducing the gap with human performance by ove
nakamods 2016/09/09
dl

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
1