Issue during inference using POTO

Megvii-BaseDetection / DeFCN

End-to-End Object Detection with Fully Convolutional Network

Apache License 2.0

495 stars 36 forks source link

Closed SunSet0864 closed 3 years ago

SunSet0864 commented 3 years ago

NMS post-processing is not used in POTO. The code: https://github.com/Megvii-BaseDetection/DeFCN/blob/a82393e290455fd11a1a088723a8050791b44c15/playground/detection/coco/poto.res50.fpn.coco.800size.3x_ms.3dmf/fcos.py#L407 says that topk_candidates and score_threshold are useless in POTO. So how to modify the codes in function "inference_single_image" when using POTO? Would you like to give some ideas? Thank you.

wjfwzzc commented 3 years ago

Just remove them and keep the sort only.