RetinaNet (Focal Loss for Dense Object Detection)

Model Zoo

Backbone	Model	imgs/GPU	lr schedule	FPS	Box AP	download	config
ResNet50-FPN	RetinaNet	2	1x	---	37.5	model	config
ResNet50-FPN	RetinaNet	2	2x	---	39.1	model	config
ResNet101-FPN	RetinaNet	2	2x	---	40.6	model	config
ResNet50-FPN	RetinaNet + FGD	2	2x	---	40.8	model	config/slim_config

Notes:

• The ResNet50-FPN are trained on COCO train2017 with 8 GPUs. Both ResNet101-FPN and ResNet50-FPN with FGD are trained on COCO train2017 with 4 GPUs.
• All above models are evaluated on val2017. Box AP=mAP(IoU=0.5:0.95).

Citation

@inproceedings{lin2017focal,
  title={Focal loss for dense object detection},
  author={Lin, Tsung-Yi and Goyal, Priya and Girshick, Ross and He, Kaiming and Doll{\'a}r, Piotr},
  booktitle={Proceedings of the IEEE international conference on computer vision},
  year={2017}
}