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 |
Tags
- deep learning
- Object Detection
- re-identification
- 딥러닝
- 3D
- center loss
- 파이썬
- 스택
- cv2
- 알고리즘
- Python
- 임계처리
- reconstruction
- Computer Vision
- transformer
- attention
- 큐
- OpenCV
- Knowledge Distillation
- numpy
- 자료구조
- NLP
- Threshold
- Deeplearning
- level2
- 논문 구현
- flame
- 프로그래머스
- point cloud
- Object Tracking
Archives
- Today
- Total
목록face recognition (1)
공돌이 공룡의 서재
![](http://i1.daumcdn.net/thumb/C150x150/?fname=https://blog.kakaocdn.net/dn/kIRMN/btrt7tTRHrl/bdbmGCKvuoNZxftndQuQK1/img.png)
A Discriminative Feature Learning Approach for Deep Face Recognition center loss 의 경우 face recognition 뿐 아니라 reid task에도 자주 적용되는 loss function으로, 최신 reID논문들을 이해하는데 필요한 loss 다. 이 함수를 처음 제시한 논문을 읽어보았다. Introduction face recognition task에서 feature는 separable + discriminative 특징이 있어야 한다. →class 간 구분이 잘 되며 특징이 뚜렷한 feature Discriminative 한 feature는 NN이나 k-NN 같은 알고리즘으로 잘 구분이 될 수 있다. 그러나 softmax loss는 f..
딥러닝/Vision
2022. 2. 23. 14:23