Marc Rußwurm
@MarcCoru
Assistant Professor at Wageningen University
20-05-2014 19:40:50
437 Tweets
1,0K Followers
737 Following
🙌⭐️ New paper by Gabeff Valentin! 🐻🌲
Vanilla CLIP (trained on natural images) is not precise when querying a camera trap dataset with text.
So, how can Vision-Language Models be adapted to wildlife images?
➡️ with WildCLIP: A specifically fine-tuned model. See paper ⬇️⬇️