Tag
NVIDIA's research team released LocateAnything, a vision-language detection model that rethinks bounding box prediction, which is trending #1 on HuggingFace.