[B! åŠæ•™å¸«ã‚ã‚Šå¦ç¿’] p_tanã®ãƒ

p_tan id:p_tan

åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã«é–¢ã™ã‚‹p_tanã®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}}

[æœ€æ–°è«–æ–‡/GAN] ãƒ©ãƒ™ãƒ«ãŒãªã„ãªã‚‰æŽ¨æ¸¬ã™ã‚Œã°ã„ã„ã€‚ - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? è«–æ–‡ç´¹ä»‹ãƒ»ç”»åƒå¼•ç”¨ Google Brainã‚ˆã‚Š2019.3.6æå‡º https://arxiv.org/pdf/1903.02271v1.pdf æœ¬ç ”ç©¶ã®GANã®ç‰¹å¾´ã¨æˆæžœ FIDã‚¹ã‚³ã‚¢ï¼ˆä½Žã„æ–¹ãŒè‰¯ã„ï¼‰ ã‚¹ã‚³ã‚¢ï¼˜ï½žï¼™ã®é–“ã«ã‚ã‚‹ç¸¦ç·šã¯ãƒ™ãƒ¼ã‚¹ãƒ©ã‚¤ãƒ³ï¼ˆã™ã¹ã¦ãƒ©ãƒ™ãƒ«ä»˜ã‘ã•ã‚ŒãŸç”»åƒã‚’ä½¿ã£ãŸBigGANï¼‰ æœ¬ç ”ç©¶ã®æ–¹æ³•ï¼ˆ$S^3GAN$ï¼‰ã§ã¯ ãƒ©ãƒ™ãƒ«ä»˜ã‘ã•ã‚ŒãŸç”»åƒã¯å…¨ä½“ã®ãŸã£ãŸï¼‘ï¼%ã«ã‚‚é–¢ã‚ã‚‰ãšSOTAã§ã‚ã‚‹BigGANã¨åŒç‰ã®æ€§èƒ½ã«ãªã£ãŸ ã¾ãŸå…¨ä½“ã®ï¼’ï¼%ã‚’ãƒ©ãƒ™ãƒ«ä»˜ã‘ã•ã‚ŒãŸç”»åƒã«ã™ã‚‹ã¨BigGANã‚’è¶…ãˆã‚‹æ€§èƒ½ã¨ãªã£ãŸ è§£åƒåº¦128Ã—128 ä¸Šæ®µ
p_tan 2019/03/12
GAN

è«–æ–‡

è§£èª¬

æ•™å¸«ãªã—å¦ç¿’

åŠæ•™å¸«ã‚ã‚Šå¦ç¿’
ãƒªãƒ³ã‚¯
ã€Šæ—¥çµŒRoboã€‹å¦ç¿’ã®æœ¬å‘½ã€ŒåŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã€ã€å°‘ãªã„æ•™å¸«ãƒ‡ãƒ¼ã‚¿ã§é«˜ã„ç²¾åº¦
æ©Ÿæ¢°å¦ç¿’ã«ãŠã„ã¦æ•™å¸«ã‚ã‚Šå¦ç¿’ã¯ã€ä¸€èˆ¬ç”»åƒèªè˜ã‚„éŸ³å£°èªè˜ãªã©ã§ç›®è¦šã¾ã—ã„æˆæžœã‚’ä¸Šã’ã¦ã„ã‚‹ã€‚ä¸€æ–¹ã§ä¸–ã®ä¸ã®ã»ã¨ã‚“ã©ã®å•é¡Œã§ã¯æ•™å¸«ãƒ‡ãƒ¼ã‚¿ã¯ã»ã¨ã‚“ã©å˜åœ¨ã›ãšã€æ•™å¸«ãªã—ãƒ‡ãƒ¼ã‚¿ãŒå¤§é‡ã«å˜åœ¨ã™ã‚‹ã€‚ã“ã®ã‚ˆã†ãªã€å¤§é‡ã®æ•™å¸«ãªã—ãƒ‡ãƒ¼ã‚¿ã¨å°‘é‡ã®æ•™å¸«ã‚ã‚Šãƒ‡ãƒ¼ã‚¿ãŒä¸Žãˆã‚‰ã‚ŒãŸå ´åˆã«å¦ç¿’ã™ã‚‹å•é¡Œã‚’ã€ŒåŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã€ã¨å‘¼ã¶ã€‚ ç”Ÿç‰©ã®ã»ã¨ã‚“ã©ã®å¦ç¿’ãŒã€ã“ã®ã‚ˆã†ãªåŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã®å•é¡Œã ã¨è€ƒãˆã‚‰ã‚Œã‚‹ã€‚äººé–“ã¯ç”Ÿã¾ã‚Œã¦ã‹ã‚‰æ§˜ã€…ãªç‰©ä½“ã‚’ã€æ£è§£ã‚’ä¸Žãˆã‚‰ã‚Œãšã«ãŸã è¦³å¯Ÿã™ã‚‹ã€‚ãã—ã¦ã‚ã‚‹æ—¥ã€ãã®ç‰©ä½“ã®ãƒ©ãƒ™ãƒ«ï¼ˆä¾‹ãˆã°ã€Œã“ã‚Œã¯è±¡ã€ã€Œã‚ã‚Œã¯ã‚ãƒªãƒ³ã€ãªã©ï¼‰ãŒä¸Žãˆã‚‰ã‚Œã‚‹ã¨ã€ãã®æ•™å¸«äº‹ä¾‹ã‚’æ•°åƒä¾‹ã‚‚ä¸Žãˆã‚‰ã‚Œãªãã¦ã‚‚é€”ç«¯ã«èªè˜ã§ãã‚‹ã‚ˆã†ã«ãªã‚‹ã€‚ã“ã†ã—ãŸç”Ÿç‰©ã®å¦ç¿’ã®éŽç¨‹ã§ã¯ã€æ•™å¸«ãªã—ã®ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰ä½•ã‚‰ã‹ã®å¦ç¿’ã‚’è¡Œã£ã¦ã„ã‚‹ã¯ãšã§ã‚ã‚‹ã€‚ åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã¯2ç¨®é¡žã«å¤§åˆ¥ åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã®æ‰‹æ³•ã¯å¤§ãã2ã¤ã«åˆ†ã‹ã‚Œã‚‹ã€‚1ã¤ç›®ã¯ã€ã€Œãƒ–ãƒ¼ãƒˆã‚¹ãƒˆãƒ©ãƒƒãƒ—æ³•ã€ã¨å‘¼ã°ã‚Œã‚‹æ–¹æ³•ã§ã‚ã‚‹ã€‚
p_tan 2017/09/26
åŠæ•™å¸«ã‚ã‚Šå¦ç¿’

è§£èª¬

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã®ãƒ¢ãƒ‡ãƒ«ä»®å®š - ã§ã‹ã„ãƒãƒ¼ã‚ºã‚’ãƒ™ãƒ¼ã‚°ãƒ«ã™ã‚‹
Machine Learning Advent Calendar 2014ã®12æ—¥ç›®ã€‚ æœ€è¿‘åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã«èˆˆå‘³ãŒã‚ã£ã¦ã¡ã‚‡ã£ã¨å‹‰å¼·ã—ã¦ã¿ãŸã®ã§ãã‚Œã«ã¤ã„ã¦æ›¸ã„ã¦ã¿ã‚‹ã€‚è‡ªåˆ†ãŒå‹‰å¼·ã—ãŸæ™‚ã«èªã‚“ã æ–‡çŒ®ã‚‚ä¸‹ã®æ–¹ã«æ›¸ã„ãŸã®ã§ã‚‚ã—èˆˆå‘³ãŒã‚ã‚Œã°ã€‚ åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã¯ãƒ©ãƒ™ãƒ«ä»˜ããƒ‡ãƒ¼ã‚¿ã«åŠ ãˆã¦ãƒ©ãƒ™ãƒ«ç„¡ã—ãƒ‡ãƒ¼ã‚¿ã‚‚ä½¿ã£ã¦å¦ç¿’ã§ãã‚‹ã¨ã„ã†ã“ã¨ã§ã™ã”ãé…åŠ›çš„ãªã‚“ã ã‘ã©ã€ä½•ã‚‚è€ƒãˆãšã«ä½¿ã†ã¨æ•™å¸«ã‚ã‚Šå¦ç¿’ã‚ˆã‚Šç²¾åº¦ãŒè½ã¡ã‚‹ã“ã¨ãŒã‚ˆãã‚ã‚‹ã€‚ ãƒ©ãƒ™ãƒ«ç„¡ã—ãƒ‡ãƒ¼ã‚¿ã¯ãã®åã®é€šã‚Šã©ã®ã‚¯ãƒ©ã‚¹ã«å±žã™ã‹ãŒåˆ†ã‹ã£ã¦ã„ãªã„ãƒ‡ãƒ¼ã‚¿ãªã®ã§ã€ä½•ã‹ã—ã‚‰ã® ãƒ¢ãƒ‡ãƒ« ã«åŸºã„ã¦ãã®ãƒ‡ãƒ¼ã‚¿ãŒã©ã®ã‚¯ãƒ©ã‚¹ã«å±žã™ã‚‹ã‹ã‚’ä»®å®šã—ã¦ã‚„ã‚‰ãªã„ã¨ã„ã‘ãªã„ã€‚ã¤ã¾ã‚Šãƒ‡ãƒ¼ã‚¿ã®åˆ†å¸ƒï¼ˆãƒ¢ãƒ‡ãƒ«ï¼‰ã«ä»®å®šã‚’ç½®ã‹ãªã„ã¨ã„ã‘ãªã„ã€‚åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ã‚’ã™ã‚‹ä¸Šã§ã¯ã“ã‚ŒãŒ ä¸€ç•ªé‡è¦ã€‚ Introduction to Semi-Supervised Learningã®p.15ã«ã‚‚ã“ã†æ›¸ã„ã¦ã‚ã‚‹ã€‚ the
p_tan 2017/03/29
æ©Ÿæ¢°å¦ç¿’

machinelearning

åŠæ•™å¸«ã‚ã‚Šå¦ç¿’

ã¾ã¨ã‚
ãƒªãƒ³ã‚¯
Iclr2016 vaeã¾ã¨ã‚
3. ICLR2016ã®ãƒˆãƒ¬ãƒ³ãƒ‰ Reinforcement Learning Unsupervised Learning Incorporating Structure Compressing Networks Incorporating Structure Initializing Networks Backprop Tricks Attention Deep Metric Learning Computer Vision Applications Visualizing Networks Do Deep Convolutional Nets Really Need to be Deep? Training-Free Methods Geometric Methods Gaussian Processes and Auto Encoders ResNet http://www.comp
p_tan 2016/10/11
VAE

deep learning

æ©Ÿæ¢°å¦ç¿’

deeplearning

ã‚¬ã‚¦ã‚¹éŽç¨‹

åŠæ•™å¸«ã‚ã‚Šå¦ç¿’
ãƒªãƒ³ã‚¯
è«–æ–‡ç´¹ä»‹ Semi-supervised Learning with Deep Generative Models
1. è«–ï¥â½‚æ–‡ç´¹ä»‹ Semi-â€â€‘â€’supervised Â Learning with Â Deep Â Generative Â Models NIPS2014èªã¿ä¼š Â @ Â æ±â¼¤å¤§, Â 2015/01/20 Preferred Â Networks, Â å¾—å±… Â èª ä¹Ÿ @beam2d 2. lï¬â€¯ ãƒ©ãƒ™ãƒ«ã‚ã‚Šãƒ‡ãƒ¼ã‚¿ãŒå°‘ãªãã¦ï¼Œãã‚Œã ã‘ã§ã¯åˆ†é›¢ï§ªæ›²â¾¯é¢ã‚’æ±ºã‚ã¥ã‚‰ã„ lï¬â€¯ ãƒ©ãƒ™ãƒ«ãªã—ãƒ‡ãƒ¼ã‚¿ã‚’ä½¿ã£ã¦ç©ºé–“ã‚’è£œé–“ã—ã¦ï¼Œã„ã„æ„Ÿã˜ã«åˆ†é›¢ï§ªæ›²â¾¯é¢ã‚’æ±ºã‚ã‚ˆã† â†’ Â åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ åŠæ•™å¸«ã‚ã‚Šå¦ç¿’ Â (semi-â€â€‘â€’supervised Â learning) 2 çŒ« â½çŠ¬ ãƒ©ãƒ™ãƒ«ã‚ã‚Šãƒ‡ãƒ¼ã‚¿ï¼ˆå°‘ãªã„ï¼‰ ãƒ©ãƒ™ãƒ«ãªã—ãƒ‡ãƒ¼ã‚¿ï¼ˆå¤šã„ï¼‰ 3. å¾“æ¥â¼¿æ‰‹æ³•ï¼šâ¼¤å¤§ãã Â 4 Â ç¨®é¡ž 3 â¾ƒè‡ªâ¼°å·±æ•™â½°ç¤ºã«ã‚ˆã‚‹å¦ç¿’ â€¢â€¯ å¦ç¿’ã—ãŸäºˆæ¸¬å™¨ã‚’ä½¿ã£ã¦ãƒ©ãƒ™ãƒ«ãªã— ãƒ‡ãƒ¼ã‚¿ã‚’ãƒ©ãƒ™ãƒ«ä»˜ã‘ã™ã‚‹ â€¢â€¯ â¼¤å¤§ãƒžãƒ¼ã‚¸ãƒ³ã®ä»®èª¬ã‚’â¼Šå…¥ã‚Œã‚‹ã“ã¨ã‚‚ã‚ ã‚‹ï¼ˆTr
p_tan 2016/10/08
deep learning

æ©Ÿæ¢°å¦ç¿’

åŠæ•™å¸«ã‚ã‚Šå¦ç¿’

ç”Ÿæˆãƒ¢ãƒ‡ãƒ«
ãƒªãƒ³ã‚¯
1