일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | ||||
4 | 5 | 6 | 7 | 8 | 9 | 10 |
11 | 12 | 13 | 14 | 15 | 16 | 17 |
18 | 19 | 20 | 21 | 22 | 23 | 24 |
25 | 26 | 27 | 28 | 29 | 30 | 31 |
Tags
- Entropy Minimization
- remixmatch paper
- mocov3
- dcgan
- GAN
- WGAN
- mme paper
- 컴퓨터구조
- UnderstandingDeepLearning
- 딥러닝손실함수
- 최린컴퓨터구조
- conjugate pseudo label paper
- ConMatch
- semi supervised learnin 가정
- adamatch paper
- Pseudo Label
- CGAN
- tent paper
- Pix2Pix
- CoMatch
- BYOL
- shrinkmatch
- CycleGAN
- dann paper
- cifar100-c
- shrinkmatch paper
- simclrv2
- Meta Pseudo Labels
- 백준 알고리즘
- SSL
Archives
- Today
- Total
목록tent paper (1)
Hello Data

논문의 풀 제목은 글 제목에 있고 주로 TENT라고 불린다. 2021 ICLR에서 억셉된 논문이고 milestone 논문 같아 읽어보았고 코드도 함께 살펴보려고 한다 https://arxiv.org/pdf/2006.10726.pdf Introduction Deep neural network는 train, test data들이 same distribution이라는 setting하에 높은 성능을 보여줬다. 그러나 이러한 성능은 두 data의 분포가 다를 경우 하락을 보여주는데 이러한 것을 dataset shift라고 한다(distribution shift라고 이해해도 문제 없을 거 같다). 모델은 unexpected weather, sensor degradation등의 variation한 상황을 맞이할 가능..
Test Time Adaption
2024. 2. 7. 17:02