Search papers, labs, and topics across Lattice.
3
0
7
Large models are emerging as a promising new paradigm for translating complex-layout document images, as shown by the ICDAR 2025 DIMT competition.
Domain-specific prompts can significantly boost document layout analysis, achieving state-of-the-art results by explicitly guiding models with dataset-aware cues.
LLMs can now excel in high-frequency decision-making tasks like UAV pursuit, thanks to a novel reward normalization and consistency loss approach that aligns global and sub-semantic policies.
