Snippets Groups Projects

Rewrite dataset evaluation (#2042)

Kai Chen authored 5 years ago

* move evaluation part to datasets

* use dataset.evaluate() in testing scripts

* remove redundant codes

* fix some log formats

* update config with evaludation

* move evaluation below dataset

* remove coco_utils.py and fix test_robustness

Co-authored-by: Jerry Jiarui XU <xvjiarui0826@gmail.com>

1d7b0a79

1d7b0a79 5 years ago

Code owners

Assign users and groups as approvers for specific file changes. Learn more.

Name	Last commit	Last update
..
README.md
ga_fast_r50_caffe_fpn_1x.py
ga_faster_r50_caffe_fpn_1x.py
ga_faster_x101_32x4d_fpn_1x.py
ga_retinanet_r50_caffe_fpn_1x.py
ga_retinanet_x101_32x4d_fpn_1x.py
ga_rpn_r101_caffe_rpn_1x.py
ga_rpn_r50_caffe_fpn_1x.py
ga_rpn_x101_32x4d_fpn_1x.py

Region Proposal by Guided Anchoring

Introduction

We provide config files to reproduce the results in the CVPR 2019 paper for Region Proposal by Guided Anchoring.

@inproceedings{wang2019region,
    title={Region Proposal by Guided Anchoring},
    author={Jiaqi Wang and Kai Chen and Shuo Yang and Chen Change Loy and Dahua Lin},
    booktitle={IEEE Conference on Computer Vision and Pattern Recognition},
    year={2019}
}

Results and Models

The results on COCO 2017 val is shown in the below table. (results on test-dev are usually slightly higher than val).

Method	Backbone	Style	Lr schd	Mem (GB)	Train time (s/iter)	Inf time (fps)	AR 1000	Download
GA-RPN	R-50-FPN	caffe	1x	5.0	0.55	13.3	68.5	model
GA-RPN	R-101-FPN	caffe	1x	7.1	0.66	9.8	69.6	model
GA-RPN	X-101-32x4d-FPN	pytorch	1x	8.5	0.88	8.5	70.0	model
GA-RPN	X-101-64x4d-FPN	pytorch	1x	11.4	1.24	6.5	70.5	model

Method	Backbone	Style	Lr schd	Mem (GB)	Train time (s/iter)	Inf time (fps)	box AP	Download
GA-Fast RCNN	R-50-FPN	caffe	1x	3.3	0.23	14.9	39.5	model
GA-Faster RCNN	R-50-FPN	caffe	1x	5.1	0.64	9.6	39.9	model
GA-Faster RCNN	R-101-FPN	caffe	1x	7.3	0.75	8.0	41.5	model
GA-Faster RCNN	X-101-32x4d-FPN	pytorch	1x	8.7	0.97	7.1	42.9	model
GA-Faster RCNN	X-101-64x4d-FPN	pytorch	1x	11.6	1.33	5.7	43.9	model
GA-RetinaNet	R-50-FPN	caffe	1x	3.2	0.50	10.7	37.0	model
GA-RetinaNet	R-101-FPN	caffe	1x	5.3	0.63	8.5	38.9	model
GA-RetinaNet	X-101-32x4d-FPN	pytorch	1x	6.7	0.87	7.5	40.3	model
GA-RetinaNet	X-101-64x4d-FPN	pytorch	1x	9.6	1.22	5.8	40.8	model

In the Guided Anchoring paper, score_thr is set to 0.001 in Fast/Faster RCNN and 0.05 in RetinaNet for both baselines and Guided Anchoring.