Classifier-Balancing

This repository contains code for the paper:

Decoupling Representation and Classifier for Long-Tailed Recognition
Bingyi Kang, Saining Xie,Marcus Rohrbach, Zhicheng Yan, Albert Gordo, Jiashi Feng, Yannis Kalantidis
[OpenReview] [Arxiv] [PDF] [Slides] [@ICLR]
Facebook AI Research, National University of Singapore
International Conference on Learning Representations (ICLR), 2020

Abstract

The long-tail distribution of the visual world poses great challenges for deep learning based classification models on how to handle the class imbalance problem. Existing solutions usually involve class-balancing strategies, e.g., by loss re-weighting, data re-sampling, or transfer learning from head- to tail-classes, but all of them adhere to the scheme of jointly learning representations and classifiers. In this work, we decouple the learning procedure into representation learning and classification, and systematically explore how different balancing strategies affect them for long-tailed recognition. The findings are surprising: (1) data imbalance might not be an issue in learning high-quality representations; (2) with representations learned with the simplest instance-balanced (natural) sampling, it is also possible to achieve strong long-tailed recognition ability with relative ease by adjusting only the classifier. We conduct extensive experiments and set new state-of-the-art performance on common long-tailed benchmarks like ImageNet-LT, Places-LT and iNaturalist, showing that it is possible to outperform carefully designed losses, sampling strategies, even complex modules with memory, by using a straightforward approach that decouples representation and classification.

 

 

If you find this code useful, consider citing our work:

@inproceedings{kang2019decoupling,
  title={Decoupling representation and classifier for long-tailed recognition},
  author={Kang, Bingyi and Xie, Saining and Rohrbach, Marcus and Yan, Zhicheng
          and Gordo, Albert and Feng, Jiashi and Kalantidis, Yannis},
  booktitle={Eighth International Conference on Learning Representations (ICLR)},
  year={2020}
}

Requirements

The code is based on https://github.com/zhmiao/OpenLongTailRecognition-OLTR.

• Python 3
• PyTorch (version >= 0.4.1)
• yaml

Dataset

• ImageNet_LT and Places_LT

  Download the ImageNet_2014 and Places_365.

• iNaturalist 2018

  • Download the dataset following here.
  • cd data/iNaturalist18, Generate image name files with this script or use the existing ones [here].

Change the data_root in main.py accordingly.

Representation Learning

1. Instance-balanced Sampling

python main.py --cfg ./config/ImageNet_LT/feat_uniform.yaml

2. Class-balanced Sampling

python main.py --cfg ./config/ImageNet_LT/feat_balance.yaml

3. Square-root Sampling

python main.py --cfg ./config/ImageNet_LT/feat_squareroot.yaml

4. Progressively-balancing Sampling

python main.py --cfg ./config/ImageNet_LT/feat_shift.yaml

Test the joint learned classifier with representation learning

python main.py --cfg ./config/ImageNet_LT/feat_uniform.yaml --test 

Classifier Learning

1. Nearest Class Mean classifier (NCM).

python main.py --cfg ./config/ImageNet_LT/feat_uniform.yaml --test --knn

2. Classifier Re-training (cRT)

python main.py --cfg ./config/ImageNet_LT/cls_crt.yaml --model_dir ./logs/ImageNet_LT/models/resnext50_uniform_e90
python main.py --cfg ./config/ImageNet_LT/cls_crt.yaml --test

3. Tau-normalization

Extract fatures

for split in train_split val test
do
  python main.py --cfg ./config/ImageNet_LT/feat_uniform.yaml --test --save_feat $split
done

Evaluation

for split in train val test
do
  python tau_norm.py --root ./logs/ImageNet_LT/models/resnext50_uniform_e90/ --type $split
done

4. Learnable weight scaling (LWS)

python main.py --cfg ./config/ImageNet_LT/cls_lws.yaml --model_dir ./logs/ImageNet_LT/models/resnext50_uniform_e90
python main.py --cfg ./config/ImageNet_LT/cls_lws.yaml --test

Results and Models

ImageNet_LT

• Representation learning

  Sampling	Many	Medium	Few	All	Model
  Instance-Balanced	65.9	37.5	7.7	44.4	ResNeXt50
  Class-Balanced	61.8	40.1	15.5	45.1	ResNeXt50
  Square-Root	64.3	41.2	17.0	46.8	ResNeXt50
  Progressively-Balanced	61.9	43.2	19.4	47.2	ResNeXt50

  For other models trained with instance-balanced (natural) sampling:
  [ResNet50] [ResNet101] [ResNet152] [ResNeXt101] [ResNeXt152]

• Classifier learning

  Classifier	Many	Medium	Few	All	Model
  Joint	65.9	37.5	7.7	44.4	ResNeXt50
  NCM	56.6	45.3	28.1	47.3	ResNeXt50
  cRT	61.8	46.2	27.4	49.6	ResNeXt50
  Tau-normalization	59.1	46.9	30.7	49.4	ResNeXt50
  LWS	60.2	47.2	30.3	49.9	ResNeXt50

iNaturalist 2018

• Representaion Learning
  We provide the pre-trained models with instance-balanced (natural) sampling for 90 and 200 epochs as follows:
  90 epochs: [ResNet50] [ResNet101] [ResNet152] 200 epochs: [ResNet50] [ResNet101] [ResNet152]
• Classifier learning
  We provide the cRT and LWS models based on a pretrained ResNet152 model (200 epochs) as follows:
  [ResNet152(cRT)] [ResNet152(LWS)]

Places_LT

• Representaion learning
  We provide a pretrained ResNet152 with instance-balanced (natural) sampling: [link]
• Classifier learning
  We provide the cRT and LWS models based on above pretrained ResNet152 model as follows:
  [ResNet152(cRT)] [ResNet152(LWS)]

To test a pretrained model:
python main.py --cfg /path/to/config/file --model_dir /path/to/model/file --test

License

This project is licensed under the license found in the LICENSE file in the root directory of this source tree (here). Portions of the source code are from the OLTR project.