Search papers, labs, and topics across Lattice.
Department of Computer Science, Tulane University
1
0
3
Boost a VLM's rare-object reasoning without any finetuning by injecting learned embeddings from vision foundation models and synonym-augmented text.
