티스토리 뷰
[2019][***]Traffic Sign Detection Using a Multi-Scale Recurrent Attention Network
Arc Lab. 2019. 8. 1. 15:181. 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 1
• Experiments on the German Traffic Sign Detection Benchmark (GTSDB) [9] and the Tsinghua-Tencent
100K (TT-100K) data set [10] show that the proposed approach is competitive when compared with state-ofthe-art approaches for traffic sign detection.
- Page 3
- Page 5
We verify our proposed approach on the GTSDB [9] and the TT-100K data set [10].
We use evaluation criteria that others have used in published research in order to compare our work on the same datasets to state-of-the-art approaches. For this reason, we use area under curve (AUC) values as the evaluation measure in the GTSDB
data set, and precision-recall for the evaluation criteria for the TT-100K data set.
- Page 6
- Page 7
- Page 8
By using information from different tasks (refer again to Fig. 8), the ’TT-100K Benchmark’ approach improves detection effectiveness, that is, the precision is 84.8% when the recall is 92.3%. ’Perceptual GAN’ raises precision by 1.0% owing to its effective generator which acquires super-resolved feature maps for things like traffic signs. Based on the SSD, ’SOS-CNN’ employs multi-scale analysis in both image and feature maps, and as a result, it effectively improves recall. Our approach also employs multi-scale analysis, and the recall and precision can be improved about 1.0% with the use of local context.
- Total
- Today
- Yesterday
- Physical Simulation
- 2D Game
- #ApacheSpark
- Library
- English
- Ragdoll
- OST
- #REST API
- ate
- 도커
- SSM
- Jekyll and Hyde
- #TensorFlow
- Meow
- Mask R-CNN
- Memorize
- Worry
- Game Engine
- aws #cloudfront
- ILoop Engine
- project
- docker
- Badge
- #ApacheZeppelin
- #ELK Stack
- some time ago
- GOD
- belief
- Sea Bottom
- sentence test
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 |