Skip to content
Code for "The more you look, the more you see: towards general object understanding through recursive refinement"
MATLAB C++ Cuda
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
+dagnn
data
functions
matconvnet @ 5053a57
.gitignore
.gitmodules
LICENSE
README.md
build_nms.m
download.m
faster_rcnn_backbone_init.m
recursive_deploy.m
recursive_get_batch_single.m
recursive_init.m
recursive_test.m
recursive_train.m
run_demo.m
run_test.m
run_train.m
setup_imdb_voc11inst.m
startup.m

README.md

The more you look, the more you see: towards general object understanding through recursive refinement

This is the source code for training and evaluating the hierarchical detection model with recursive refinement module.

The code is tested with MATLAB R2016b. If you run into any problems, please file an issue, or contact Jingyan Wang.

Installation

  1. Clone the repo:
# use the --recursive option to clone the MatConvNet submodule
git clone --recursive https://github.com/jingyanw/recursive-refinement.git
  1. Build the matconvnet directory. Follow the MatConvNet installation instructions.

  2. Build the NMS module (same as Faster-RCNN): run build_nms.m

Downloads

Run download.m to download the following data to $ROOT/data:

  • Pre-trained ImageNet VGG-16 model:

pretrained/imagenet-vgg-verdeep-16.mat

  • The cluster subclasses with using the IOU metric on the object masks. The IOU between two shape clusters within the same class is roughly 0.75:

clusters/clusters-shape-thresh25.mat

  • The raw data and devkit for PASCA-VOC2012, with SBD augmentation:

    VOCdevkit/VOC2012
    VOCdevkit/VOCcode
    VOCdevkit/VOC-SBD
    
  • A trained VGG-16 model:

    models/shape-thresh25-vgg16-epoch7.mat

Demo

Run run_demo.m. The demo performs detection and instance segmentation on an example image.

Training

Run run_train.m to train the model with the default setting. This approximately takes 2 days on a single GPU. Multi-GPU support is untested.

Testing

Run run_test.m to test the provided off-the-shelf model, with the following performance on detection and instance segmentation:

Detection

AP (@0.5) AP (@0.7)
69.0 45.8

Instance segmentation

AP^r (@0.5) AP^r (@0.7)
54.6 19.2
You can’t perform that action at this time.