Skip to content
RRPN in pytorch 1.0 ----- Arbitrary-Oriented Scene Text Detection via Rotation Proposals
Branch: master
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.
configs test config May 3, 2019
demo results May 3, 2019
docker first commit May 3, 2019
maskrcnn_benchmark Update May 14, 2019
rotation first commit May 3, 2019
tests first commit May 3, 2019
tools first commit May 3, 2019
.flake8 first commit May 3, 2019
.gitignore first commit May 3, 2019 first commit May 3, 2019
LICENSE first commit May 3, 2019 first commit May 3, 2019 Update May 14, 2019 & May 9, 2019 & May 9, 2019


RRPN in pytorch, which is implemented into facebook's benchmark: Its caffe version can be viewed at:

alt text


  • From original repo: In pytorch 1.0, Somehow faster than original repo in both training and inference.
  • Training and evaluation checked: Testing in IC15 with training data in {IC13, IC15, IC17mlt}, and receives Fscore of 83% vs. 81% in caffe repo.
  • What's new: RRoI Pooling is replaced with RRoI Align(bilinear interpolation for sampling), FPN structure supported, easy to change various backbones for different purposes.


Check for installation instructions.

Configuring your dataset

  • Your dataset path can be set in $RRPN_ROOT/maskrcnn_benchmark/config/ We implemented interface for {IC13, IC15, IC17mlt, LSVT, ArT} for common use(Start from line 96):
 "RRPN_train": {  # including IC13 and IC15
                # 'IC13': 'Your dataset path',
            "split": 'train'
  • Add your dataset? You need to form a dict array as follows:
im_info = {
    'gt_classes': your class_id array,
    'max_classes': your class_id array,
    'image': path to access one image,
    'boxes': rotate box in {cx, cy, w, h, θ},
    'flipped': Not supported, just False, 
    'gt_overlaps': overlaps fill with 1 (gt with gt),
    'seg_areas': H * W for an rbox,
    'height': height of an image,
    'width': width of an image,
    'max_overlaps': overlaps fill with 1 (gt with gt),
    'rotated': just True

Examples can be seen in $RRPN_ROOT/maskrcnn_benchmark/data/ Your data API should be add to the variable DATASET:

    'Your Dataset Name': 'Your Dataset API'


# create your data cache directory
mkdir data_cache
# In your root of RRPN
python tools/ --config-file=configs/rrpn/e2e_rrpn_R_50_C4_1x_ICDAR13_15_17_trial.yaml
  • Multi-GPU phase is not testing yet, be careful to use GPU more than 1.


  • Using $RRPN_ROOT/demo/ to test images you want. The demo will generate a text for your detected coodinates.
  • Showing the detected image by ture the variable vis to True.


  • Enjoy it with all the codes.
  • Citing us if you find it work in your projects.
    author = {Jianqi Ma},
    title = {{RRPN in pytorch}},
    year = {2019},
    howpublished = {\url{}},
    Author = {Jianqi Ma and Weiyuan Shao and Hao Ye and Li Wang and Hong Wang and Yingbin Zheng and Xiangyang Xue},
    Title = {Arbitrary-Oriented Scene Text Detection via Rotation Proposals},
    journal = {IEEE Transactions on Multimedia},
You can’t perform that action at this time.