GC-Net

Weizeng Lu, Xi Jia, Weicheng Xie, Linlin Shen, Yicong Zhou, and Jinming Duan: Geometry Constrained Weakly Supervised Object Localization, In the European Conference on Computer Vision (ECCV) 2020

Overview of GC-Net

We propose a geometry constrained network, termed GC-Net, for weakly supervised object localization (WSOL). GC-Net consists of three modules: a detector, a generator and a classiﬁer. The detector predicts the object location represented by a set of coeﬃcients describing a geometric shape (i.e. ellipse or rectangle), which is geometrically constrained by the mask produced by the generator. The classiﬁer takes the resulting masked images as input and performs two complementary classiﬁcation tasks (object and background). To make the mask more compact and more complete, we propose a novel multi-task loss function that takes into account area of the geometric shape, the categorical crossentropy and the negative entropy. In contrast to previous approaches, GC-Net is trained end-to-end and predict object location without any post-processing (e.g. thresholding) that may require additional tuning. Extensive experiments on the CUB-200-2011 and ILSVRC2012 datasets show that GC-Net outperforms state-of-the-art methods by a large margin.

Setup

Clone this repo:

git clone git@github.com:lwzeng/GC-Net.git

Create an Anaconda environment:

python==3.6
PyTorch==1.1.0

Download the dataset and modify the data path in the code.

Training

1. Training the classifier
cd classifier
python classifier_vgg.py or python classifier_google.py

2. Training the generator
cd generator
python train_generator.py

3. Training the detector
cd detector
python detector.py

Evaluation

cub200:

cd detector
python test.py

imagenet:

cd detector
python test_imagenet.py

Visualization