search

lucazanella / lavad

Official implementation of "Harnessing Large Language Models for Training-free Video Anomaly Detection", CVPR 2024
https://lucazanella.github.io/lavad/
32 stars 0 forks source link

Congrats on the nice work! #2

Closed wyzjack closed 2 months ago

wyzjack commented 2 months ago

Congrats on the nice work!

I am writing to ask about when you are planning to release your code? Thanks so much.

Also, is it possible to share your extracted CLIP features on UCF Crime and XD-Violence dataset? Thanks in advance.

I would appreciate it very much if you could reply.

lucazanella commented 2 months ago

Hi, thank you for your interest in our work!

I take this opportunity to apologise to everyone who is waiting for the code to be released. I currently have very limited bandwidth due to an impending deadline. I will release the code by the end of this month at the latest. I will also release the data, including the captions, the temporal summaries, and the indexes containing their text embeddings. For this work, I did not store the pre-extracted image/video features, but if you are interested in the CLIP features for UCF-Crime and XD-Violence you can find them in this other repository. However, these features come from a CLIP ViT/B, whereas I used ImageBind's image and video encoder for this work.

wyzjack commented 2 months ago

Great, thanks so much for letting me know!