FreeAnchor

Model Name:*

Description with Markdown (optional):

# FreeAnchor: Learning to Match Anchors for Visual Object Detection

## Introduction

[ALGORITHM]

```latex
@inproceedings{zhang2019freeanchor,
  title   =  {{FreeAnchor}: Learning to Match Anchors for Visual Object Detection},
  author  =  {Zhang, Xiaosong and Wan, Fang and Liu, Chang and Ji, Rongrong and Ye, Qixiang},
  booktitle =  {Neural Information Processing Systems},
  year    =  {2019}
}
```

## Results and Models

| Backbone | Style   | Lr schd | Mem (GB) | Inf time (fps) | box AP | Config | Download |
|:--------:|:-------:|:-------:|:--------:|:--------------:|:------:|:------:|:--------:|
| R-50     | pytorch | 1x      | 4.9      | 18.4 | 38.7 | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/free_anchor/retinanet_free_anchor_r50_fpn_1x_coco.py) | [model](http://download.openmmlab.com/mmdetection/v2.0/free_anchor/retinanet_free_anchor_r50_fpn_1x_coco/retinanet_free_anchor_r50_fpn_1x_coco_20200130-0f67375f.pth) &#124; [log](http://download.openmmlab.com/mmdetection/v2.0/free_anchor/retinanet_free_anchor_r50_fpn_1x_coco/retinanet_free_anchor_r50_fpn_1x_coco_20200130_095625.log.json) |
| R-101       | pytorch | 1x   | 6.8      | 14.9 | 40.3 | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/free_anchor/retinanet_free_anchor_r101_fpn_1x_coco.py) | [model](http://download.openmmlab.com/mmdetection/v2.0/free_anchor/retinanet_free_anchor_r101_fpn_1x_coco/retinanet_free_anchor_r101_fpn_1x_coco_20200130-358324e6.pth) &#124; [log](http://download.openmmlab.com/mmdetection/v2.0/free_anchor/retinanet_free_anchor_r101_fpn_1x_coco/retinanet_free_anchor_r101_fpn_1x_coco_20200130_100723.log.json) |
| X-101-32x4d | pytorch | 1x   | 8.1      | 11.1 | 41.9 | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/free_anchor/retinanet_free_anchor_x101_32x4d_fpn_1x_coco.py) | [model](http://download.openmmlab.com/mmdetection/v2.0/free_anchor/retinanet_free_anchor_x101_32x4d_fpn_1x_coco/retinanet_free_anchor_x101_32x4d_fpn_1x_coco_20200130-d4846968.pth) &#124; [log](http://download.openmmlab.com/mmdetection/v2.0/free_anchor/retinanet_free_anchor_x101_32x4d_fpn_1x_coco/retinanet_free_anchor_x101_32x4d_fpn_1x_coco_20200130_095627.log.json) |

**Notes:**

- We use 8 GPUs with 2 images/GPU.
- For more settings and models, please refer to the [official repo](https://github.com/zhangxiaosong18/FreeAnchor).

Paper:*

Code URL (optional):

lr sched	1x
Memory (M)	6800.0
Backbone Layers	101
inference time (s/im)	0.06711

lr sched	1x
Memory (M)	4900.0
Backbone Layers	50
inference time (s/im)	0.05435

lr sched	1x
Memory (M)	8100.0
Backbone Layers	101
inference time (s/im)	0.09009

BENCHMARK	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
COCO minival	FreeAnchor (X-101-32x4d, 1x, pytorch)	box AP	41.9	# 56
COCO minival	FreeAnchor (R-101, 1x, pytorch)	box AP	40.3	# 71
COCO minival	FreeAnchor (R-50, 1x, pytorch)	box AP	38.7	# 87

open-mmlab / mmdetection

FreeAnchor: Learning to Match Anchors for Visual Object Detection

Introduction

Results and Models

Results

Object Detection on COCO minival

Object Detection

Architecture	FreeAnchor, ResNet, Focal Loss
lr sched	1x
Memory (M)	6800.0
Backbone Layers	101
inference time (s/im)	0.06711
SHOW MORE
SHOW LESS

Architecture	ResNeXt, FreeAnchor, Focal Loss
lr sched	1x
Memory (M)	8100.0
Backbone Layers	101
inference time (s/im)	0.09009
SHOW MORE
SHOW LESS

Backbone	Style	Lr schd	Mem (GB)	Inf time (fps)	box AP	Config	Download
R-50	pytorch	1x	4.9	18.4	38.7	config	model \| log
R-101	pytorch	1x	6.8	14.9	40.3	config	model \| log
X-101-32x4d	pytorch	1x	8.1	11.1	41.9	config	model \| log