<aside> 👋🏻
This fork would aim to provide a reliable, up-to-date, and easy-to-use version of YOLOX that maintains its Apache License, ensuring it remains accessible for academic and commercial use. The primary focus is on fixing compatibility issues and maintaining core functionality, with additional features added based on future success.
The project will require ongoing maintenance to keep up with Python and PyTorch updates, but the level of effort should be manageable with focused work on dependency management rather than major code changes.
</aside>
YOLOX is an object detection framework that was last updated in 2022 and is now facing compatibility issues with current Python environments, dependencies, and platforms like Google Colab. There is community demand for a maintained, Apache-licensed alternative to other YOLO implementations (e.g. YOLOV…)
The fork would live under this organization: https://github.com/pixeltable at *https://github.com/pixeltable/YOLOX.* At Pixeltable, we are a well-funded, and a small team with decades of experiences in managing open source infrastructure projects and frameworks from Apache Parquet, Apache Impala, internal Machine Learning frameworks and services at Amazon, Airbnb, Twitter, YellowBrick, MapR, Dremio and more. We have no intentions in monetizing YOLOX as it’s just another library that we integrate with for our users. Here’s an example of how it’s being used in Pixeltable: