Search papers, labs, and topics across Lattice.
2
901
5
7
Ditch slow, token-by-token box generation: LocateAnything's Parallel Box Decoding (PBD) boosts VLM grounding speed and accuracy by decoding entire bounding boxes at once.
Open-source multimodal models just leveled up: InternVL3 rivals closed-source titans like GPT-4o by pre-training vision and language together from the start.
