Search papers, labs, and topics across Lattice.
1
0
3
5
Injecting LLM-generated text descriptions into a vision model boosts UI element detection, particularly for ambiguous cases where visual cues alone fall short.
