티스토리 뷰
[2018][***]Small Object Detection Using Deep Feature Pyramid Networks
Arc Lab. 2019. 7. 31. 11:151. 인용 논문
Traffic-sign detection and classification in the wild
Zhe Zhu, Dun Liang, Songhai Zhang, Xiaolei Huang, Baoli Li, Shimin Hu;
The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 2110-2118
2. 인용 부분
- Page 4-5
To detect small objects from a high-resolution image, we design a two-stage architecture as shown in Fig. 3. For region proposal stage, two top-down feature pyramids are built to obtain the high-level semantic feature maps at all scales, which boost the discriminative power of feature representation for small object.
For classification stage, we adopt the idea of Densenet [6]. To improve accuracy, focal loss [14] is used to supervise the proposal network.
Focal loss [14] achieves great success on solving the problem of class imbalance during training.
- Page 7
- Page 8
- Total
- Today
- Yesterday
- Memorize
- belief
- some time ago
- #TensorFlow
- Badge
- #ApacheZeppelin
- Jekyll and Hyde
- OST
- Library
- Mask R-CNN
- aws #cloudfront
- English
- #ELK Stack
- Physical Simulation
- sentence test
- docker
- Ragdoll
- Meow
- 2D Game
- Game Engine
- Worry
- Sea Bottom
- GOD
- ate
- SSM
- #REST API
- #ApacheSpark
- 도커
- project
- ILoop Engine
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 |