detectron/configs/Detectron1-Comparisons

• Comparisons:

Detectron2's default settings and a few implementation details are different from Detectron.

The differences in implementation details are shared in
Compatibility with Other Libraries.

The differences in default config includes:

• Use scale augmentation during training.
• Use L1 loss instead of smooth L1 loss.
• Use POOLER_SAMPLING_RATIO=0 instead of 2.
• Use ROIAlignV2.

In this directory, we provide a few configs that mimic Detectron's behavior as close as possible.
This provides a fair comparison of accuracy and speed against Detectron.

Name	lr sched	train time (s/iter)	inference time (s/im)	train mem (GB)	box AP	mask AP	kp. AP	model id	download
Faster R-CNN	1x	0.219	0.048	3.1	36.9			137781054	model | metrics
Keypoint R-CNN	1x	0.313	0.082	5.0	53.1		64.2	137781195	model | metrics
Mask R-CNN	1x	0.273	0.052	3.4	37.8	34.9		137781281	model | metrics

Comparisons:

• Faster R-CNN: Detectron's AP is 36.7, similar to ours.
• Keypoint R-CNN: Detectron's AP is box 53.6, keypoint 64.2. Fixing a Detectron's
  bug lead to a drop in box AP, and can be
  compensated back by some parameter tuning.
• Mask R-CNN: Detectron's AP is box 37.7, mask 33.9. We're 1 AP better in mask AP, due to more correct implementation.

For speed comparison, see benchmarks.