๐Ÿค– ์ธ๊ณต์ง€๋Šฅ

์ƒ์œ„ ํ† ํ”ฝ์œผ๋กœ ๋ชจ์•„์ง„ ๋‰ด์Šค ๊ธฐ์‚ฌ, ๊ทธ๋ฆฌ๊ณ  ๊ธฐํƒ€ ์ •๋ณด๋ฅผ ๋‹ด์€ ๊ธ€์„, ํ•˜์œ„ ํ† ํ”ฝ๋ณ„๋กœ ์žฌ๋ถ„๋ฅ˜ํ•˜๋Š” ์—ฐ๊ตฌ๋ฅผ ์ง„ํ–‰ํ•˜๋ฉฐ ๋Œ€๋Ÿ‰์˜ ๋ฌธ์„œ ๋ฐ์ดํ„ฐ ๋‚ด์—์„œ unsupervisedํ•˜๊ฒŒ ํ† ํ”ฝ์„ ์ถ”์ถœํ•˜๊ณ  ๋™์‹œ์— ํด๋Ÿฌ์Šคํ„ฐ๋ง๊นŒ์ง€ ์ง„ํ–‰ํ•˜๋Š” ๊ฒƒ์„ Clustering Topic Modeling์ด๋ผ๊ณ  ํ•œ๋‹ค. 1. Topic Modeling์˜ ํ•œ๊ณ„ ๊ทธ๋Ÿฐ๋ฐ, ํŠนํžˆ ๋‰ด์Šค์™€ ๊ฐ™์ด ์‚ฌ๊ฑด์„ ๋‹ค๋ฃฌ ๊ธ€๋“ค์„ ํด๋Ÿฌ์Šคํ„ฐ๋งํ•˜๋Š” ๊ณผ์ •์—์„œ ๊ธฐ์กด Topic Modeling๋งŒ์œผ๋กœ๋Š” ํ•œ๊ณ„๋ฅผ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค. ํ˜„์žฌ Benchmark๋กœ ์‚ผ๊ณ  ์žˆ๋Š” Top2Vec์˜ ๊ฒฝ์šฐ๋ฅผ ์‚ดํŽด๋ณด๋ฉด, https://github.com/ddangelov/Top2Vec GitHub - ddangelov/Top2Vec: Top2Vec learns jointly embedded topic, document..
๋Œ€๋Ÿ‰์˜ ํ…์ŠคํŠธ ๋ฐ์ดํ„ฐ๋ฅผ ํด๋Ÿฌ์Šคํ„ฐ๋งํ•˜๋Š” ์—ฐ๊ตฌ๋ฅผ ํ•˜๊ณ  ์žˆ๋‹ค. ๊ธฐ์กด์—๋Š” K-means๋กœ, ๊ฐ€์ง€๊ณ  ์žˆ๋Š” ์ปดํ“จํŒ… ์ž์›์ด ๋งŽ์ง€ ์•Š์•„ ๊ฐ€์žฅ ๊ฐ€๋ฒผ์šด ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์‚ฌ์šฉํ•˜๋‹ค๊ฐ€ HDBSCAN์„ ์ ‘ํ•˜๊ณ  ๊ดœ์ฐฎ๋‹ค๊ณ  ์ƒ๊ฐํ•˜์—ฌ ๋น„๊ตํ•˜๋Š” ๊ธ€์„ ์ž‘์„ฑํ•œ๋‹ค. 1. ์ •์˜ K-means: https://ko.wikipedia.org/wiki/K-%ED%8F%89%EA%B7%A0_%EC%95%8C%EA%B3%A0%EB%A6%AC%EC%A6%98 k-ํ‰๊ท  ์•Œ๊ณ ๋ฆฌ์ฆ˜ - ์œ„ํ‚ค๋ฐฑ๊ณผ, ์šฐ๋ฆฌ ๋ชจ๋‘์˜ ๋ฐฑ๊ณผ์‚ฌ์ „ ์œ„ํ‚ค๋ฐฑ๊ณผ, ์šฐ๋ฆฌ ๋ชจ๋‘์˜ ๋ฐฑ๊ณผ์‚ฌ์ „. k-ํ‰๊ท  ์•Œ๊ณ ๋ฆฌ์ฆ˜(K-means clustering algorithm)์€ ์ฃผ์–ด์ง„ ๋ฐ์ดํ„ฐ๋ฅผ k๊ฐœ์˜ ํด๋Ÿฌ์Šคํ„ฐ๋กœ ๋ฌถ๋Š” ์•Œ๊ณ ๋ฆฌ์ฆ˜์œผ๋กœ, ๊ฐ ํด๋Ÿฌ์Šคํ„ฐ์™€ ๊ฑฐ๋ฆฌ ์ฐจ์ด์˜ ๋ถ„์‚ฐ์„ ์ตœ์†Œํ™”ํ•˜๋Š” ๋ฐฉ์‹์œผ๋กœ ๋™์ž‘ ko.wikipedia.org K..
์ด ํฌ์ŠคํŠธ๋Š” Attention Mechanism๋ฟ๋งŒ ์•„๋‹ˆ๋ผ ๊ทธ ์ „์˜ ๊ธฐ๋ณธ์ ์ธ ์š”์†Œ์— ๋Œ€ํ•ด์„œ๋„ ํ—ท๊ฐˆ๋ฆฌ๋Š” ๋ถ„๋“ค์„ ์œ„ํ•ด ์ž‘์„ฑ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. "Attention is All You Need"๋Š” Transformer ์•„ํ‚คํ…์ฒ˜๋ฅผ ์ฒ˜์Œ ์ œ์‹œํ•œ ๋…ผ๋ฌธ์ธ๋ฐ์š”. ๋ณธ ์•„ํ‚คํ…์ฒ˜ ๋‚ด์˜ ๊ธฐ๋ณธ์ ์ธ ์š”์†Œ๋ถ€ํ„ฐ ๋‹ค์†Œ ๊ณผํ•˜๊ฒŒ ์ •๋ฆฌํ•  ๊ฒƒ์ด๊ธฐ ๋•Œ๋ฌธ์— ๋Œ€๋žต 3๊ฐœ์˜ ํฌ์ŠคํŒ… ์—ฐ์žฌ๋ฅผ ํ†ตํ•ด ์ด๋ฅผ ์„ค๋ช…ํ•˜๊ณ ์ž ํ•ฉ๋‹ˆ๋‹ค. Attention Is All You Need The dominant sequence transduction models are based on complex recurrent or convolutional neural networks in an encoder-decoder configuration. The best performing mode..
๋…ผ๋ฌธ์˜ ๊ฐœ์š” - ์ง€๋‚œ ๊ธ€ ์ฐธ๊ณ  [CVPR 22'] Dressing in the Wild by Watching Dance Video - 1 https://arxiv.org/abs/2203.15320 Dressing in the Wild by Watching Dance Videos While significant progress has been made in garment transfer, one of the most applicable directions of human-centric image generation, existing works overlook the in-the-wild im inseon.tistory.com 1. Conditional Person Segmentation ๊ธฐ์กด์˜ pose t..
https://arxiv.org/abs/2203.15320 Dressing in the Wild by Watching Dance Videos While significant progress has been made in garment transfer, one of the most applicable directions of human-centric image generation, existing works overlook the in-the-wild imagery, presenting severe garment-person misalignment as well as noticeable degr arxiv.org Dressing in the Wild by Watching Dance Video๋Š” ํ‹ฑํ†ก์„ ์„œ๋น„..
inthree3
'๐Ÿค– ์ธ๊ณต์ง€๋Šฅ' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๊ธ€ ๋ชฉ๋ก (2 Page)