NSFW JS for Web Apps

NSFW JS is a JavaScript library designed for identifying inappropriate images in real-time on user devices. Utilizing TensorFlowJS, this library achieves a notable accuracy rate of 93% for detecting specific image content. It features CameraBlur Protection, which automatically blurs any flagged images, enhancing user safety. Regular updates and the inclusion of new models ensure that the library remains effective and relevant. The library is available under the MIT license, supporting open-source collaboration.

This tool is particularly beneficial for developers looking to implement content filtering directly within web applications. As a lightweight solution at just 4.2MB, NSFW JS facilitates client-side image detection without requiring server-side processing. It also includes a mobile demo for testing on various devices, making it accessible for developers on the go. Community contributions are encouraged, fostering continuous improvement and expansion of its capabilities.