diff --git a/README.md b/README.md index e8d45af64fa3b0dcc942a1b54a1e494fba1d8473..5935a541a3ea1fd3c8b2ae436f7d5c29c367265b 100644 --- a/README.md +++ b/README.md @@ -46,7 +46,7 @@ Please refer to [changelog.md](docs/changelog.md) for details and release histor ## Benchmark and model zoo Supported methods and backbones are shown in the below table. -Results and models are available in the [Model zoo](docs/MODEL_ZOO.md). +Results and models are available in the [model zoo](docs/MODEL_ZOO.md). | | ResNet | ResNeXt | SENet | VGG | HRNet | |--------------------|:--------:|:--------:|:--------:|:--------:|:-----:| @@ -71,18 +71,22 @@ Results and models are available in the [Model zoo](docs/MODEL_ZOO.md). | FreeAnchor | 鉁� | 鉁� | 鈽� | 鉁� | 鉁� | | NAS-FPN | 鉁� | 鉁� | 鈽� | 鉁� | 鉁� | | ATSS | 鉁� | 鉁� | 鈽� | 鉁� | 鉁� | +| FSAF | 鉁� | 鉁� | 鈽� | 鉁� | 鉁� | +| PAFPN | 鉁� | 鉁� | 鈽� | 鉁� | 鉁� | Other features - [x] [CARAFE](configs/carafe/README.md) - [x] [DCNv2](configs/dcn/README.md) - [x] [Group Normalization](configs/gn/README.md) - [x] [Weight Standardization](configs/gn+ws/README.md) -- [x] [OHEM](configs/faster_rcnn_ohem_r50_fpn_1x_coco.py) -- [x] Soft-NMS +- [x] [OHEM](configs/faster_rcnn/faster_rcnn_r50_fpn_ohem_1x_coco.py) +- [x] [Soft-NMS](configs/faster_rcnn/faster_rcnn_r50_fpn_soft_nms_1x_coco.py) - [x] [Generalized Attention](configs/empirical_attention/README.md) - [x] [GCNet](configs/gcnet/README.md) -- [x] [Mixed Precision (FP16) Training](configs/fp16) +- [x] [Mixed Precision (FP16) Training](configs/fp16/README.md) - [x] [InstaBoost](configs/instaboost/README.md) +- [x] [FSAF](configs/fsaf/README.md) +- [x] [PAFPN](configs/pafpn/README.md) ## Installation diff --git a/configs/faster_rcnn/faster_rcnn_r50_fpn_soft_nms_1x_coco.py b/configs/faster_rcnn/faster_rcnn_r50_fpn_soft_nms_1x_coco.py new file mode 100644 index 0000000000000000000000000000000000000000..05f61483e4b3dbb6f3469e72fada40cb97881dff --- /dev/null +++ b/configs/faster_rcnn/faster_rcnn_r50_fpn_soft_nms_1x_coco.py @@ -0,0 +1,11 @@ +_base_ = [ + '../_base_/models/faster_rcnn_r50_fpn.py', + '../_base_/datasets/coco_detection.py', + '../_base_/schedules/schedule_1x.py', '../_base_/default_runtime.py' +] + +test_cfg = dict( + rcnn=dict( + score_thr=0.05, + nms=dict(type='soft_nms', iou_thr=0.5), + max_per_img=100)) diff --git a/docs/model_zoo.md b/docs/model_zoo.md index fb0f38a08dd59b12f7e42151d028c74a74b1747f..57c76815c2f79b6d856cccee218f27293f08bde9 100644 --- a/docs/model_zoo.md +++ b/docs/model_zoo.md @@ -127,6 +127,19 @@ Please refer to [ATSS](https://github.com/open-mmlab/mmdetection/blob/master/con We also benchmark some methods on [PASCAL VOC](https://github.com/open-mmlab/mmdetection/blob/master/configs/pascal_voc), [Cityscapes](https://github.com/open-mmlab/mmdetection/blob/master/configs/cityscapes) and [WIDER FACE](https://github.com/open-mmlab/mmdetection/blob/master/configs/wider_face). +## Speed benchmark +We compare the training speed of Mask R-CNN with some other popular frameworks (The data is copied from [detectron2](https://github.com/facebookresearch/detectron2/blob/master/docs/notes/benchmarks.md)). + +| Implementation | Throughput (img/s) | +|----------------------|--------------------| +| [Detectron2](https://github.com/facebookresearch/detectron2) | 61 | +| [MMDetection](https://github.com/open-mmlab/mmdetection) | 60 | +| [maskrcnn-benchmark](https://github.com/facebookresearch/maskrcnn-benchmark/) | 51 | +| [tensorpack](https://github.com/tensorpack/tensorpack/tree/master/examples/FasterRCNN) | 50 | +| [simpledet](https://github.com/TuSimple/simpledet/) | 39 | +| [Detectron](https://github.com/facebookresearch/Detectron) | 19 | +| [matterport/Mask_RCNN](https://github.com/matterport/Mask_RCNN/) | 14 | + ## Comparison with Detectron2 We compare mmdetection with [Detectron2](https://github.com/facebookresearch/detectron2.git) in terms of speed and performance. diff --git a/requirements/build.txt b/requirements/build.txt index 84c500a07f493a6f6bf1dee4585c264467cae861..0da9929b1cdcdc5f5942c10c0093ff017b29519c 100644 --- a/requirements/build.txt +++ b/requirements/build.txt @@ -1,3 +1,3 @@ # These must be installed before building mmdetection numpy -torch>=1.1 +torch>=1.3 diff --git a/requirements/runtime.txt b/requirements/runtime.txt index 13194bca0e83e3d32e663684267bc032503deaff..8b3b12af2a5d393e60a9e63c731b3d6c7bc02cfe 100644 --- a/requirements/runtime.txt +++ b/requirements/runtime.txt @@ -1,9 +1,9 @@ matplotlib -mmcv>=0.5.0 +mmcv>=0.5.1 numpy # need older pillow until torchvision is fixed Pillow<=6.2.2 six terminaltables -torch>=1.1 +torch>=1.3 torchvision