๐Ÿ›๏ธ 2024.01.03

์‚ฌ์ดํด ๋ฐ ํ—ฌ์Šค์˜ ์—ฌํŒŒ๋กœ 9์‹œ์— ์ž ์„ ์ฒญํ•ด๋ฒ„๋ ธ๋‹ค.
๋Œ€๋žต ์ƒˆ๋ฒฝ 4์‹œ์ •๋„์— ๊ธฐ์ƒํ•  ๊ฒƒ์œผ๋กœ ๊ธฐ๋Œ€ํ•œ ๋‚ด ๋งˆ์Œ๊ณผ๋Š” ๋‹ฌ๋ฆฌ 11์‹œ์— ์นผ๊ฐ™์ด ๊ธฐ์ƒํ•ด ๋ฒ„๋ ธ๋‹ค...
์ง„์งœ ๋งํ–ˆ๋‹คใ… ใ… 
์ถœ๊ทผ์ค€๋น„์‹œ๊ฐ„์ด 8์‹œ๊ฐ„์ด๋‚˜ ๋‚จ์•„๋ฒ„๋ฆฐ๊ฒƒ...
 
์–ด์ฐŒํ•  ๋ฐฉ๋„ ์—†์ด ๋ฐ”๋กœ ๊ณต๋ถ€ ์‹œ์ž‘!๐Ÿ”ฅ
 
๊ณต๋ถ€๋ฅผ ํ•˜๋‹ค๋ณด๋‹ˆ ์–ด๋Š๋ง 4์‹œ... ์ถœ๊ทผ์‹œ๊ฐ„๊นŒ์ง€ 2์‹œ๊ฐ„ ๋‚จ์•„ ์ž ์„ ์ž˜๊นŒ ๋ง๊นŒ ๊ณ ๋ฏผํ–ˆ์ง€๋งŒ ๋ฐ”๋กœ ๋ƒ…๋‹ค ์ž๋ฒ„๋ ธ๋‹ค.
 
 
 
 
 
 
 
 
 
 

๐Ÿข ETRI (2024.01.04)

๐Ÿš• ๊ฐ€๋Š” ๊ธธ

๐Ÿƒ๐Ÿป ์ถœ๋ฐœ!

์ƒˆ๋ฒฝ 6์‹œ์— ์•Œ๋žŒ์„ ๋งž์ท„์ง€๋งŒ ๊ท€์‹ ๊ฐ™์ด ์ผ์–ด๋‚œ ์‹œ๊ฐ์€ 8์‹œ 10๋ถ„!!

๋ฌผ๋ก  ์ผ์–ด๋‚  ๋•Œ๋„ ๊ท€์‹ ๊ฐ™์ด "์œผ์•…! ๋Šฆ์—ˆ๋‹ค, ์น˜์ฝ”์ฟ  ์น˜์ฝ”์ฟ (์ง€๊ฐ)~"ํ•˜๋ฉด์„œ ์ผ์–ด๋‚˜ ๋ฒ„๋ ธ๋‹ค.
(๋˜ ๋Šฆ์ž  ์ž๋ฒ„๋ ธ๋‹คใ… ใ…  ๋ฐ์ž๋ทด๊ฐ€...)

 
์šด์ด ์ข‹๊ฒŒ๋„(?) ํƒ€์Šˆ(๐Ÿšฒ)๊ฐ€ ๋‚  ๋ฐ˜๊ฒจ์ฃผ๊ณ  ์žˆ์„ ์‹œ๊ฐ„ ๋”ฐ์œˆ ์—†์—ˆ๋‹ค!
๋ฐ”๋กœ ์นด์นด์˜คT ํƒ์‹œ ์žก๊ณ  ์ถœ๋ฐœํ–ˆ๋‹ค!!!
 
 

๐Ÿข ETRI ๋„์ฐฉ!

ํƒ์‹œ์—์„œ ๋‚ด๋ฆฌ๋‹ˆ 8์‹œ 50๋ถ„๊ฒฝ... 9์‹œ๊นŒ์ง€์ธ์ค„ ์•Œ๊ณ  ํ—ˆ๊ฒ์ง€๊ฒ ๋‹ฌ๋ฆฌ๋‹ˆ ์ •๋ฌธ์—์„œ ๋ณด์•ˆ์š”์›?๋“ค์ด ์–ด๋””๊ฐ€์‹œ๋Š”๊ฑฐ๋ƒ ๋ฌผ์—ˆ๊ณ , ์ธํ„ดํ•˜๋Ÿฌ ๊ฐ„๋‹ค๊ณ  ๋‹ตํ–ˆ๋‹ค.
๋‹ค์‹œ ์ฐพ์•„๋ณด๋‹ˆ 9์‹œ 30๋ถ„๊นŒ์ง€์˜€๋˜๊ฑด ์•ˆ๋น„๋ฐ€ใ… ใ…  ๋‚ด 7500์›๐Ÿ’ฆ

 
์–ด์ฐŒ์ €์ฐŒ ํ†ต๊ณผํ•œ ๋’ค ์•„๋ž˜ ์‚ฌ์ง„์˜ ํฐ ๊ฑด๋ฌผ๋กœ ๋“ค์–ด๊ฐ€์„œ OT๋ฅผ ๋“ค์—ˆ๋‹ค.

 

๊ฐ€๊นŒ์ด ๊ฐ€๋ณด๋‹ˆ 1๋™ ๊ฑด๋ฌผ์ด๋ผ๊ณ  ํ•˜๋Š”๋ฐ ์—ฌ๊ธฐ ํšŒ์˜์‹ค์—์„œ ๊ฐ„๋‹จํ•œ? OT๋ฅผ ๊ฐ€์กŒ๋‹ค.
OT์—์„œ ๋ณด์•ˆ์„œ์•ฝ์„œ ์“ฐ๊ณ  ์ถœ์ž…์ฆ์„ ๋ฐ›์•˜๋‹คใ…Žใ…Ž

 

๐Ÿข ๊ณผ์—ฐ ๋‚ด๊ฐ€ ์žˆ์„ ๊ฑด๋ฌผ์€?

OT๊ฐ€ ๋๋‚˜๊ณ  12๋™์— ๋ฐฐ์ •๋ฐ›์•„ ๊ฐ€๋Š”๊ธธ์— ์—ฐ๋ชป๋„ ๋ณด์—ฌ์„œ ๊ตฌ๊ฒฝํ•˜๋‹ค๋ณด๋‹ˆ

 
 
12๋™์— ๋„์ฐฉํ•ด๋ฒ„๋ ธ๋‹ค!
12๋™ ๊ฑด๋ฌผ์ด ์ œ์ผ ๋ฉ‹์ง„๋“ฏ?

๋„์ฐฉํ•ด์„œ ๊ฐ„๋‹จํ•˜๊ฒŒ ์ž๋ฆฌ๋ฐฐ์ • ๋ฐ ํ”„๋กœ์ ํŠธ ์ฃผ์ œ๋ฐฐ๊ฒฝ์„ค๋ช…๋งŒ ๋“ฃ๊ณ  ์งˆ์˜์‘๋‹ต์„ ์ง„ํ–‰ํ–ˆ๋‹ค.
ํ”„๋กœ์ ํŠธ ์ฃผ์ œ๋„ ๋น„๋ฐ€...
 
 

๐Ÿš ๋ฐฅ์ด๋‹ค ๋ฐฅ!

๋„์ฐฉํ•ด์„œ ๊ฐ„๋‹จํ•˜๊ฒŒ ์ž๋ฆฌ๋ฐฐ์ • ๋ฐ ํ”„๋กœ์ ํŠธ ์ฃผ์ œ๋ฐฐ๊ฒฝ์„ค๋ช…๋งŒ ๋“ฃ๊ณ  ์งˆ์˜์‘๋‹ต ํ›„ ๋ฐฅ์„ ๋จน์œผ๋Ÿฌ ๊ฐ”๋‹ค.

์„œํŽธ์ œ์™€ ๋™ํŽธ์ œ(ํŠน์‹)๊ฐ€ ์žˆ์—ˆ๋Š”๋ฐ, ๋™ํŽธ์ œ๊ฐ€ 2500์› ๋” ๋น„์ŒŒ๋‹ค.
์‹ ๊ธฐํ–ˆ๋˜ ์ ์€ ๋ชจ๋“  ๋ฐฅ๊ณผ ๋ฐ˜์ฐฌ์„ ์–‘๊ป ๋‹ด์„ ์ˆ˜ ์žˆ๋‹ค๋Š” ์ ์ด์—ˆ๋‹ค! ๊ทธ๋ž˜์„œ ์œ„ ์‚ฌ์ง„์ฒ˜๋Ÿผ ์ง„์งœ ๋งŽ์ด ๋ฐ›์•„์„œ ๋จน์—ˆ๋‹ค.
 
 
 
 
 

๐Ÿ“– ๊ณต๋ถ€! (13:00-18:00)

์‹์‚ฌ ํ›„ ๊ต์ˆ˜๋‹˜๊ป˜์„œ ์นœ์ ˆํ•˜๊ฒŒ ETRI ๊ฑด๋ฌผ ์•ˆ๋‚ดํ•ด ์ฃผ์…จ๊ณ , ๊ฐ„๋‹จํ•˜๊ฒŒ ๋ฐ์ดํŠธ?๋ฅผ ์ฆ๊ธด๋‹ค์Œ ๊ณต๋ถ€์— ๋Œ์ž…ํ–ˆ๋‹ค.
๊ณต๋ถ€๋‚ด์šฉ์€ Domain Adaptation!
๊ณต๋ถ€๋‚ด์šฉ์€ ETRI(๊ณต๋ถ€ ์ฐธ์กฐ!)

 

[DA]: Visual Domain Adaptation

Part 1. Basic Concepts. &. Traditional Methods ๐Ÿ“Œ Basic Concepts ๐Ÿค”Domain๊ณผ DA๋ž€? Domain: ๋ฐ์ดํ„ฐ๊ฐ€ ์œ ๋ž˜ํ•œ ๊ณณ์ด๋‚˜ ํŠน์ • ๋ถ„ํฌ๋ฅผ ์˜๋ฏธ DA ๊ธฐ๋ณธ๊ฐœ๋…: ์•„๋ž˜์™€ ๊ฐ™์ด source์™€ target ๋ถ„ํฌ๊ฐ„์˜ Domain Shift๊ฐ€ ๋ฐœ์ƒํ•˜๊ณ , ๊ทธ ์ •

chan4im.tistory.com

 
 
 
 
 
 
 

๐Ÿƒ๐Ÿป ํ‡ด๊ทผ! (18:00 -)

์•„ ์ง„์งœ ํž˜๋“ค๋‹คใ…‹ใ…‹ใ„ฒ
ํ‡ด๊ทผ๊ธธ ํ•œ์žฅ  ์ฐฐ์นต


๋นจ๋ฆฌ๊ฐ€์„œ ๋ฒค์น˜๋ž‘ ์Šค์ฟผํŠธ๋‚˜ ํ•ด์•ผ๊ฒ ๋‹คใ…‹ใ…‹
 
 
 
 

๐Ÿ“Œ TODO List:

1. Detection๋ถ„์•ผ Domain Adaptation ๋…ผ๋ฌธ ์ฝ๊ธฐ
  i) Deep Visual Domain Adaptation - VI.C. Object Detection part
 ii) Domain Adaptive Faster R-CNN for Object Detection in the Wild

2. PaperswithCode Search

3. ๊ด€๋ จ GitHub Search
 

Deep Visual Domain Adaptation: A Survey

Deep domain adaption has emerged as a new learning technique to address the lack of massive amounts of labeled data. Compared to conventional methods, which learn shared feature subspaces or reuse important source instances with shallow representations, de

arxiv.org

 

Domain Adaptive Faster R-CNN for Object Detection in the Wild

Object detection typically assumes that training and test data are drawn from an identical distribution, which, however, does not always hold in practice. Such a distribution mismatch will lead to a significant performance drop. In this work, we aim to imp

arxiv.org

 

Papers with Code - Paper tables with annotated results for Unsupervised Domain Adaptation of Object Detectors: A Survey

Paper tables with annotated results for Unsupervised Domain Adaptation of Object Detectors: A Survey

paperswithcode.com


 

GitHub - zhaoxin94/awesome-domain-adaptation: A collection of AWESOME things about domian adaptation

A collection of AWESOME things about domian adaptation - GitHub - zhaoxin94/awesome-domain-adaptation: A collection of AWESOME things about domian adaptation

github.com

 

Papers with Code - Domain Adaptation

**Domain Adaptation** is the task of adapting models across domains. This is motivated by the challenge where the test and training datasets fall from different data distributions due to some factor. Domain adaptation aims to build machine learning models

paperswithcode.com

 

Papers with Code - Open World Object Detection

Open World Object Detection is a computer vision problem where a model is tasked to: 1) identify objects that have not been introduced to it as `unknown', without explicit supervision to do so, and 2) incrementally learn these identified unknown categories

paperswithcode.com

 
 

'2024 winter > ETRI(์ผ์ƒ)' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๋‹ค๋ฅธ ๊ธ€

[ETRI]2024.01.10  (4) 2024.01.10
[ETRI]2024.01.09  (2) 2024.01.09
[ETRI]2024.01.08  (4) 2024.01.08
[ETRI]2024.01.05  (4) 2024.01.05
[ETRI]2024.01.03  (4) 2024.01.03

+ Recent posts