MaochengHu 576cda45b8 first commit 2 anni fa
..
README.md 576cda45b8 first commit 2 anni fa
cascade_rcnn_dcn_r50_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
cascade_rcnn_dcn_x101_vd_64x4d_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
faster_rcnn_dcn_r101_vd_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
faster_rcnn_dcn_r50_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
faster_rcnn_dcn_r50_vd_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
faster_rcnn_dcn_r50_vd_fpn_2x_coco.yml 576cda45b8 first commit 2 anni fa
faster_rcnn_dcn_x101_vd_64x4d_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
mask_rcnn_dcn_r101_vd_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
mask_rcnn_dcn_r50_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa
mask_rcnn_dcn_r50_vd_fpn_2x_coco.yml 576cda45b8 first commit 2 anni fa
mask_rcnn_dcn_x101_vd_64x4d_fpn_1x_coco.yml 576cda45b8 first commit 2 anni fa

README.md

Deformable ConvNets v2

骨架网络 网络类型 卷积 每张GPU图片个数 学习率策略 推理时间(fps) Box AP Mask AP 下载 配置文件
ResNet50-FPN Faster c3-c5 1 1x - 42.1 - 下载链接 配置文件
ResNet50-vd-FPN Faster c3-c5 1 1x - 42.7 - 下载链接 配置文件
ResNet50-vd-FPN Faster c3-c5 1 2x - 43.7 - 下载链接 配置文件
ResNet101-vd-FPN Faster c3-c5 1 1x - 45.1 - 下载链接 配置文件
ResNeXt101-vd-FPN Faster c3-c5 1 1x - 46.5 - 下载链接 配置文件
ResNet50-FPN Mask c3-c5 1 1x - 42.7 38.4 下载链接 配置文件
ResNet50-vd-FPN Mask c3-c5 1 2x - 44.6 39.8 下载链接 配置文件
ResNet101-vd-FPN Mask c3-c5 1 1x - 45.6 40.6 下载链接 配置文件
ResNeXt101-vd-FPN Mask c3-c5 1 1x - 47.3 42.0 下载链接 配置文件
ResNet50-FPN Cascade Faster c3-c5 1 1x - 42.1 - 下载链接 配置文件
ResNeXt101-vd-FPN Cascade Faster c3-c5 1 1x - 48.8 - 下载链接 配置文件

注意事项:

  • Deformable卷积网络v2(dcn_v2)参考自论文Deformable ConvNets v2.
  • c3-c5意思是在resnet模块的3到5阶段增加dcn.

Citations

@inproceedings{dai2017deformable,
  title={Deformable Convolutional Networks},
  author={Dai, Jifeng and Qi, Haozhi and Xiong, Yuwen and Li, Yi and Zhang, Guodong and Hu, Han and Wei, Yichen},
  booktitle={Proceedings of the IEEE international conference on computer vision},
  year={2017}
}
@article{zhu2018deformable,
  title={Deformable ConvNets v2: More Deformable, Better Results},
  author={Zhu, Xizhou and Hu, Han and Lin, Stephen and Dai, Jifeng},
  journal={arXiv preprint arXiv:1811.11168},
  year={2018}
}