GAIA-cv

Introduction

GAIA-cv is a fundamental vision library for customized AI solutions, and supports many projects as follows:

• GAIA-det: GAIA toolbox that provides automatic AI solutions for object detection.
• GAIA-ssl: GAIA toolbox that provides automatic AI solutions for self-supervised learning.
• GAIA-seg: GAIA toolbox that provides automatic AI solutions for semantic segmentation. Almost ready, coming soon!

It provides functionalities that help the customization of AI solutions.

• Design and apply customized search space of any type with little efforts.
• Manage models in search space according to your rules.
• Integrate datasets of various sources.

Requirements

• Python 3.6+
• CUDA 10.0+
• mmcv >= 1.2.7 & mmcv < 1.3.0
• Others (See requirements.txt)

Installation

• Install the full version of mmcv. Please refer to this.
• Install gaiavision in three lines:

git clone https://github.com/GAIA-vision/GAIA-cv . && cd GAIA-cv
pip install -r requirements.txt
pip install -e .

Usage

import gaiavision # load registered modules

Acknowledgements

• This repo is constructed on top of the awesome toolbox mmcv.
• Some of the ideas are inspired by an amazing work named BigNAS.

Contributors

We encourage every practitioner in the community to contribute to GAIA-vision. The coutributions including but not limited to implementing new search spaces, new architectures, customized solutions on public datasets and etc. are welcomed. For now, the contributors include Junran Peng, Xingyuan Bu, Qing Chang, Haoran Yin.