Notice
Recent Posts
Recent Comments
Link
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
- Bidirectional Transformers for Language Understanding
- Transformer
- contrastive learning
- machine unlearning
- Swin Transformer
- FLOPS
- 인공지능
- DeepLearning
- Semantic Segmentation
- ECCV
- self supervised learning
- 딥러닝
- GAN # Label Smoothing
- multi head attention
- mordernize
- Attention
- convnext
- class activation mapping
- boundary unlearning
- Instance segmentation
- CBAM
- BERT
- Deepfake
- ICML
- ConvNet
- You Only Look Once #YOLO
- simclr
- NT-Xent(InFoNCE)
- NLP
- Convolution Block Attention Module
Archives
- Today
- Total
목록mordernize (1)
륵늘

이 논문을 간단요약하면, Transformer의 기능들을 CNN에 하나씩 단계적으로 접목시키는 논문으로 FackBook에서 만든 논문으로 2020년 CVPR에 Accept 된 논문이고, Citation의 수는 6월 10일 기준으로 4344회이다. 2020년 VIT는 함께 기존의 State-of-the-art였던 ConvNeXt 기반의 모델들을 뛰어넘었다.VIT는 일반적인 Computer vision task에 적용하기 힘들다는 어려움이 있었는데, ConvNet의 Sliding window strategy를 통해서 Transformer도 Computer vision 분야의 Backbone으로 사용 가능하다는 것을 보이면서 많은 Vision 분야에서 좋은 성능을 보여주게 된다. 하지만 Tranformer..
딥러닝
2024. 7. 29. 17:44