https://github.com/airctic/icevision

An Agnostic Computer Vision Framework - Pluggable to any Training Library: Fastai, Pytorch-Lightning with more to come
https://github.com/airctic/icevision

ai annotation-parsers coco-dataset coco-parser computer-vision deep-learning effecientdet fastai faster-rcnn mask-rcnn object-detection pycocotools python pytorch pytorch-lightning tutorials voc-dataset voc-parser

Last synced: 2 days ago
JSON representation

An Agnostic Computer Vision Framework - Pluggable to any Training Library: Fastai, Pytorch-Lightning with more to come

• Host: GitHub
• URL: https://github.com/airctic/icevision
• Owner: airctic
• License: apache-2.0
• Created: 2020-05-04T01:57:02.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2024-11-18T13:53:14.000Z (3 months ago)
• Last Synced: 2024-11-19T09:52:26.606Z (2 months ago)
• Topics: ai, annotation-parsers, coco-dataset, coco-parser, computer-vision, deep-learning, effecientdet, fastai, faster-rcnn, mask-rcnn, object-detection, pycocotools, python, pytorch, pytorch-lightning, tutorials, voc-dataset, voc-parser
• Language: Python
• Homepage: https://airctic.github.io/icevision/
• Size: 848 MB
• Stars: 848
• Watchers: 24
• Forks: 150
• Open Issues: 83
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE
  • Code of conduct: CODE_OF_CONDUCT.md

Awesome Lists containing this project

• awesome-fastai - IceVision - an agnostic computer vision framework, pluggable to any training library (Libraries / Computer Vision)

README

        


  

  
An Agnostic Computer Vision Framework



* * * * *

    

[![tests](https://github.com/airctic/icevision/workflows/tests/badge.svg?event=push)](https://github.com/airctic/icevision/actions?query=workflow%3Atests)

[![docs](https://github.com/airctic/icevision/workflows/docs/badge.svg)](https://airctic.com)

[![codecov](https://codecov.io/gh/airctic/icevision/branch/master/graph/badge.svg)](https://codecov.io/gh/airctic/icevision)

[![PyPI version](https://badge.fury.io/py/icevision.svg)](https://badge.fury.io/py/icevision)

[![Downloads](https://pepy.tech/badge/icevision)](https://pepy.tech/project/icevision)

[![black](https://img.shields.io/badge/code%20style-black-000000.svg)](https://github.com/psf/black)

[![license](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://github.com/airctic/icevision/blob/master/LICENSE)

[![Discord](https://img.shields.io/discord/735877944085446747?label=Discord&logo=Discord)](https://discord.gg/2jqrwrQ)

* * * * *

IceVision is the first agnostic computer vision framework to offer a curated collection with hundreds of high-quality pre-trained models from [Torchvision](https://github.com/pytorch/vision), Open MMLab's [MMDetection](https://github.com/open-mmlab/mmdetection), Ultralytic's [YOLOv5](https://github.com/ultralytics/yolov5), Ross Wightman's [EfficientDet](https://github.com/rwightman/efficientdet-pytorch) and soon PyTorch Image Models. It orchestrates the end-to-end deep learning workflow allowing to train networks with easy-to-use robust high-performance libraries such as [PyTorch-Lightning](https://github.com/PyTorchLightning/pytorch-lightning) and [Fastai](https://github.com/fastai/fastai).

**IceVision Unique Features:**

- Data curation/cleaning with auto-fix

- Access to an exploratory data analysis dashboard

- Pluggable transforms for better model generalization 

- Access to hundreds of neural net models

- Access to multiple training loop libraries

- Multi-task training to efficiently combine object detection, segmentation, and classification models 

## Installation

```bash

pip install icevision[all]

```

For more installation options, check our [docs](https://airctic.com/0.7.0/install/).

**Important:** We currently only support Linux/MacOS.

## Quick Example: How to train the **Fridge Objects Dataset**



![image](images/icevision-readme.png)

![image](images/icevision-end-to-end-training.gif)

## Happy Learning!

If you need any assistance, feel free to:

[Join our Forum](https://discord.gg/JDBeZYK)