๐Ÿข ETRI (2024.01.09)

๐Ÿ“Œ 01.08-TODO List:

1. ๋…ผ๋ฌธ List๋งˆ์ € ์ฝ๊ธฐ
โˆ™ Multi-Granularity Alignment Domain Adaptation for Object Detection (CVPR 2022)
โˆ™ Cross-domain adaptive teacher for object detection (CVPR 2022)
โˆ™ ConfMix: Unsupervised Domain Adaptation for Object Detection via Confidence-based Mixing (WACV 2023)
โˆ™ Object detection based on semi-supervised domain adaptation for imbalanced domain resources (Machine Vision and Applications 31 2020)

2. Deep Learning 2024(Bishop): Chapter 2 ์ฝ๊ณ  ๊ณต๋ถ€ํ•˜๊ธฐ
โˆ™ 2.1  The Rule of Probability
โˆ™ 2.2  Probability Densities
โˆ™ 2.3  The Gaussian Distribution
โˆ™ 2.4  Transformation of Densities
โˆ™ 2.5  Information Theory
โˆ™ 2.6. Bayesian Probabilites

3. ์ฃผ๋ง์— ๊ณต๋ถ€ํ•  ๊ฒƒ: WanDB์‚ฌ์šฉ๋ฒ• ์ œ๋Œ€๋กœ ์•Œ๊ธฐ!
i) WanDB๋ž€?
ii) Sweep์•Œ๊ธฐโ—๏ธ
iii) WanDB Visualization Method (feat. Confusion Matrix)

4. ๋…ผ๋ฌธ review: A Free Lunch for Unsupervised Domain Adaptive Object Detection without Source Data
โˆ™ Abstract
โˆ™ Introduction (- ํ™”)
โˆ™ Related Works (๋‚ด์šฉ ์ ์Œ - ์ˆ˜)
โˆ™ Source free Domain Adaptive Object Detection (๋‚ด์šฉ ์งฑ๋งŽ์Œ - ์ˆ˜)
โˆ™ Experiments (2๋ฒˆ์งธ๋กœ ๋งŽ์Œ - ๋ชฉ)
โˆ™ Conclusion
 

A Free Lunch for Unsupervised Domain Adaptive Object Detection without Source Data | Proceedings of the AAAI Conference o

 

ojs.aaai.org

 
 
 

๐Ÿš ์ถœ๊ทผ! (8:10 - 8:20)

6์‹œ ์•Œ๋žŒ์ด ์ง€๋…ํ•˜๊ฒŒ๋„ ๋‚  ๊นจ์›Œ์„œ ๊ฒจ์šฐ๊ฒจ์šฐ ์…”ํ‹€๋ฒ„์Šค ์‹œ๊ฐ„์— ๋งž์ถฐ ์ผ์–ด๋‚ฌ๋‹ค...
๊ทธ๋ž˜๋„ ๋ฐฅ์€ ๋ง›์žˆ๊ฒŒ ์•ผ๋ฌด์ง€๊ฒŒ ๋จน๊ณ ์™”๋‹ค ํžˆํžˆ๐Ÿคฃ
์ถœํ‡ด๊ทผ์€ ์…”ํ‹€๋ฒ„์Šค๊ฐ€ ์—ญ์‰ฌ ์ œ์ผ ๋น ๋ฅธ๊ฑฐ ๊ฐ™๋‹ค. (Shortest Path Algorithm?)
 
 
 
 
 

๐Ÿ“– ๊ณต๋ถ€! (8:30-12:00)

์…”ํ‹€๋ฒ„์Šค ์šดํ–‰ํ•˜๋Š” ๊ณณ์ด ๋‚ด ์ž์ทจ๋ฐฉ ๊ทผ์ฒ˜๋ผ๋Š” ๊ฒƒ์„ ์•Œ๊ณ ? ์ผ์ฐ์™€์„œ ๊ณต๋ถ€๋ฅผ ์‹œ์ž‘ํ•  ์ˆ˜ ์žˆ์—ˆ๋‹ค.
์ง€๋‚œ์ฃผ์— ๋ฏธ๋ฆฌ ์ •๋ฆฌํ•œ ๋…ผ๋ฌธ ๋ฆฌ์ŠคํŠธ์— ๋Œ€ํ•œ Abstract์™€ Conclusion์— ๋Œ€ํ•ด ์ •๋ฆฌํ•ด ๋ณด์•˜๋‹ค.
(๋‚จ์€ ๋…ผ๋ฌธ 3ํŽธ ์ •๋ฆฌ์™„๋ฃŒ. 8:30~10:10)
https://chan4im.tistory.com/220

 

[DA]: Relative Research paper Brief.

Survey Paper โˆ™ Deep Domain Adaptive Object Detection: a Survey (IEEE 2020) โˆ™ Unsupervised Domain Adaptation of Object Detectors: A Survey (IEEE 2023) Conference Paper โˆ™ Domain Adaptive Faster R-CNN for Object Detection in the Wild (CVPR 2018) โˆ™ Abs

chan4im.tistory.com

 
 
 
 
 
 
 
 
 
 

๐Ÿš ๋ฐฅ์ด๋‹ค ๋ฐฅ! (12:30~13:00)

๋งˆ์ฐฌ๊ฐ€์ง€๋กœ ๋™ํŽธ์ œ(ํ•œ์‹)
๋‹ญ๊ฐˆ๋น„๊ฐ€ ๋‚˜์™”๋Š”๋ฐ ์‚ฌ์ง„ ์ง€๊ธˆ๋ณด๋‹ˆ ์ง„์งœ ๋งŽ์ด๋„ ๋‹ด์•˜๋‹คใ…‹ใ…‹ใ…‹
(์‚ด๋บ€๋‹ค๊ณ  ์‹์ด์กฐ์ ˆํ•˜๊ณ  ์žˆ๋Š”๋ฐ ์ด๊ฑฐ๋•œ์‹œ ๋ฐ”๋กœ ๋ผ์ง€๐Ÿท๋ ๋“ฏใ…‹ใ…‹ใ…‹)


 
 
 
 
 
 

 
 

๐Ÿ“– ๊ณต๋ถ€! (13:30-18:00)

WandB ์ •๋ฆฌ๋ฅผ ๋ชจ๋‘ ๋๋ƒˆ๋‹ค.

 

'Deep Learning : Vision System/MLOps Tool: Weights & Biass' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๊ธ€ ๋ชฉ๋ก

์ปดํ“จํ„ฐ๊ณผํ•™, ์ธ๊ณต์ง€๋Šฅ์— ๋Œ€ํ•œ ๊ณต๋ถ€๋ฅผ ์—ด์‹ฌํžˆ ํ•˜๋Š” ๊ณต๋Œ€์ƒ

chan4im.tistory.com

 
์ฃผ๋ง์— ๋”ฅ๋Ÿฌ๋‹๊ณผ ๋…ผ๋ฌธ๋ฆฌ๋ทฐ์— ์ข€ ๋” ์ง‘์ค‘ํ•ด์„œ ๊ณต๋ถ€ํ•  ์ˆ˜ ์žˆ์„ ๊ฒƒ ๊ฐ™๋‹ค ํžˆํžˆ๐Ÿ˜œ.

๋…ผ๋ฌธ ์ดˆ๋ก,์„œ๋ก ,๊ฒฐ๋ก ํŒŒํŠธ๋Š” ๋‹ค ์ฝ์—ˆ๋‹ค
ํ›„ํ›„..
 
 

 

๐Ÿš ํ‡ด๊ทผ! (18:30 - 19:00)

ํ‡ด๊ทผ์ด ์ œ์ผ ์ข‹๋‹ค๋Š” ์ง์žฅ์ธ์˜ ๋ง์— ๊ณต๊ฐ์ด ์ ์ ๋œ๋‹คใ…‹ใ…‹ใ…‹
์•„์นจ์ถœ๊ทผ๋•Œ๋ฌธ์— ๊ทธ๋ƒฅ ์กธ๋ฆฌ๋‹คใ…‹ใ…‹ ํ”ผ๊ณค๐Ÿฉธํ•˜๊ณ ใ… ใ… 


 
 
 
 

๐Ÿ‹๐Ÿป ์šด๋™! (19:00 -)

๋ฒค์น˜ํ•˜๊ณ  ์ง‘๊ฐ€๋Š”๋ฐ ํญ์„ค์ด...
์–ด์šฐ...
 
 

 
 
 

๐Ÿ“Œ TODO List:

1. ๋…ผ๋ฌธ review: A Free Lunch for Unsupervised Domain Adaptive Object Detection without Source Data
โˆ™ Related Works (๋‚ด์šฉ ์ ์Œ - ์ˆ˜)
โˆ™ Source free Domain Adaptive Object Detection (๋‚ด์šฉ ์งฑ๋งŽ์Œ - ์ˆ˜)
โˆ™ Experiments (2๋ฒˆ์งธ๋กœ ๋งŽ์Œ - ๋ชฉ)


2. ์ฃผ์ค‘์— ํ•  ๊ฒƒ: Deep Learning 2024(Bishop)-Chapter 2 ์ฝ๊ณ  ๊ณต๋ถ€ํ•˜๊ธฐ
โˆ™ 2.1
โˆ™ 2.2
โˆ™ 2.3
โˆ™ 2.4
โˆ™ 2.5
โˆ™ 2.6
 

A Free Lunch for Unsupervised Domain Adaptive Object Detection without Source Data | Proceedings of the AAAI Conference o

 

ojs.aaai.org

 

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

[ETRI]2024.01.11  (0) 2024.01.11
[ETRI]2024.01.10  (4) 2024.01.10
[ETRI]2024.01.08  (4) 2024.01.08
[ETRI]2024.01.05  (4) 2024.01.05
[ETRI]2024.01.04  (0) 2024.01.04

+ Recent posts