Code for paper [ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot] (ICCV 2021, oral))

Last update: May 09, 2022

Related tags

Overview

ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot

This repository is the official PyTorch implementation of ICCV-21 paper ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot.

Prerequirements

To install the environment.

conda env create -f environment.yml
conda activate ACE
git clone https://github.com/NVIDIA/apex
cd apex
pip install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./

Data format

The annotation of a dataset is a dict consisting of two field: annotations and num_classes. The field annotations is a list of dict with image_id, fpath, im_height, im_width and category_id.

Here is an example.

{
    'annotations': [
                    {
                        'image_id': 1,
                        'fpath': '/data/iNat18/images/train_val2018/Plantae/7477/3b60c9486db1d2ee875f11a669fbde4a.jpg',
                        'im_height': 600,
                        'im_width': 800,
                        'category_id': 7477
                    },
                    ...
                   ]
    'num_classes': 8142
}

Usage

Training

#bash data_parallel_train.sh configuration_file_path GPU_indexes
bash data_parallel_train.sh configs/cifar100_im100.yaml 0,1

Testing

#python valid.py configuration_file_path
python valid.py configs/cifar100_im100.yaml

Trained models

Acknowledgement

This project is developed based on Bag of tricks @AAAI-21, thanks for their works!

Citation

@inproceedings{cai2021ace,
  title={ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot},
  author={Cai, Jiarui and Wang, Yizhou and Hwang, Jenq-Neng},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
  pages={112--121},
  year={2021}
}

Code for paper [ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot] (ICCV 2021, oral))

Related tags

Overview

ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot

Prerequirements

Data format

Usage

Training

Testing

Acknowledgement

Citation

Owner

Jiarui

TEA: A Sequential Recommendation Framework via Temporally Evolving Aggregations

ivadomed is an integrated framework for medical image analysis with deep learning.

Doing the asl sign language classification on static images using graph neural networks.

Codebase for testing whether hidden states of neural networks encode discrete structures.

A library of extension and helper modules for Python's data analysis and machine learning libraries.

A curated list of awesome resources combining Transformers with Neural Architecture Search

CMUA-Watermark: A Cross-Model Universal Adversarial Watermark for Combating Deepfakes (AAAI2022)

Code for SALT: Stackelberg Adversarial Regularization, EMNLP 2021.

A curated list of awesome game datasets, and tools to artificial intelligence in games

A method to perform unsupervised cross-region adaptation of crop classifiers trained with satellite image time series.

A curated list of awesome deep long-tailed learning resources.

Assessing the Influence of Models on the Performance of Reinforcement Learning Algorithms applied on Continuous Control Tasks

Cross-lingual Transfer for Speech Processing using Acoustic Language Similarity

Official code for NeurIPS 2021 paper "Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GAN"

A multilingual version of MS MARCO passage ranking dataset

Smart edu-autobooking - Johnson @ DMI-UNICT study room self-booking system

Code for Generating Disentangled Arguments with Prompts: A Simple Event Extraction Framework that Works

From Fidelity to Perceptual Quality: A Semi-Supervised Approach for Low-Light Image Enhancement (CVPR'2020)

In the AI for TSP competition we try to solve optimization problems using machine learning.

(NeurIPS 2021) Realistic Evaluation of Transductive Few-Shot Learning