pout = 1), each of which contributesnequivariantscalars. The function\(u({r}_{ij}):{\mathbb{R}}\to {\mathbb{R}}\)is again the smooth cutoff envelope from Eq. (7). The purpose of the latent MLP is to compress and integrate...
We further proposed an efficient fisheye person detection model that is equipped with a rotation-equivariant training strat- egy. The physical locations of detected persons are formu- lated based on the fisheye model and the altitude of the cam- era. We e...
Transformation-equivariant 3d ob- ject detection for autonomous driving. arXiv preprint arXiv:2211.11962, 2022. 3, 6 [58] Yizheng Wu, Min Shi, Shuaiyuan Du, Hao Lu, Zhiguo Cao, and Weicai Zhong. 3d instances as 1d kernels. In ECCV, pages 235–252. Springer,...
Redet: A rotation-equivariant detector for aerial object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, TN, USA, 20–25 June 2021; pp. 2786–2795. [Google Scholar] Li, W.; Chen, Y.; Hu, K.; Zhu, J. Oriented RepPoints for ...