OpenGVLab / InternImage

[CVPR 2023 Highlight] InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions
https://arxiv.org/abs/2211.05778
MIT License
2.52k stars 234 forks source link

请问在offset计算过程中加入注意力机制能否加强卷积的偏好? #170

Open ID-Mott opened 1 year ago