Detecting Rotated Objects as Gaussian Distributions and Its 3-D Generalization
| Autor: | Xue Yang, Gefan Zhang, Xiaojiang Yang, Yue Zhou, Wentao Wang, Jin Tang, Tao He, Junchi Yan |
|---|---|
| Rok vydání: | 2022 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Artificial Intelligence (cs.AI) Computational Theory and Mathematics Computer Science - Artificial Intelligence Artificial Intelligence Computer Vision and Pattern Recognition (cs.CV) Applied Mathematics Computer Science - Computer Vision and Pattern Recognition Computer Vision and Pattern Recognition Software Machine Learning (cs.LG) |
| Zdroj: | IEEE Transactions on Pattern Analysis and Machine Intelligence. :1-18 |
| ISSN: | 1939-3539 0162-8828 |
| DOI: | 10.1109/tpami.2022.3197152 |
| Popis: | Existing detection methods commonly use a parameterized bounding box (BBox) to model and detect (horizontal) objects and an additional rotation angle parameter is used for rotated objects. We argue that such a mechanism has fundamental limitations in building an effective regression loss for rotation detection, especially for high-precision detection with high IoU (e.g. 0.75). Instead, we propose to model the rotated objects as Gaussian distributions. A direct advantage is that our new regression loss regarding the distance between two Gaussians e.g. Kullback-Leibler Divergence (KLD), can well align the actual detection performance metric, which is not well addressed in existing methods. Moreover, the two bottlenecks i.e. boundary discontinuity and square-like problem also disappear. We also propose an efficient Gaussian metric-based label assignment strategy to further boost the performance. Interestingly, by analyzing the BBox parameters' gradients under our Gaussian-based KLD loss, we show that these parameters are dynamically updated with interpretable physical meaning, which help explain the effectiveness of our approach, especially for high-precision detection. We extend our approach from 2-D to 3-D with a tailored algorithm design to handle the heading estimation, and experimental results on twelve public datasets (2-D/3-D, aerial/text/face images) with various base detectors show its superiority. Comment: 19 pages, 11 figures, 16 tables, accepted by TPAMI 2022. Journal extension for GWD (ICML'21) and KLD (NeurIPS'21). arXiv admin note: text overlap with arXiv:2101.11952 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|