MELD Logo

Multi-animal Ensemble Landmark Detector

Multi-animal Ensemble Landmark Detector (MELD) is an AI-driven tool for accurately detecting and tracking animal facial landmarks. Currently, it supports cats' and dogs' facial landmark schemas grounded in AnimalFACS.

Let's try Your Video


Notice For better performance, make sure that the animal’s face is clearly visible on most frames.


Notice Animal identification is in development. Currently only videos with one animal are supported.







25% 50% 75% 100%



? By default, the video will be saved on the server.