EdgeTAM is a Hugging Face Space that enables users to perform object tracking in videos by selecting positive and negative points on the first frame. After uploading a video, users mark points to include or exclude, and the model tracks the target object across all subsequent frames. It is a demonstration of the EdgeTAM model originally from Meta.
EdgeTAM is an Image generation project. It focuses on tracking specific objects through video sequences by simply selecting points on the first frame. EdgeTAM is a consumer product aimed at computer vision researchers and developers. It is available for free. It ships for the web, and it can be self-hosted.
It is developed by merve, and it first shipped in 2024. Among its 4 catalogued features are Point-based Tracking, Object Inclusion/Exclusion, and Video Upload.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do