Search papers, labs, and topics across Lattice.
Affiliation:
4
0
6
The latest AI City Challenge reveals that integrating foundation models with synthetic data design can significantly boost performance in complex urban scenarios.
Internal dual modality prompting can significantly enhance spatial detail while preserving spectral fidelity in hyperspectral pansharpening.
Hallucination rates in MLLMs can be slashed by up to 60% through targeted interventions in attention dynamics, revealing a critical pathway to improving model reliability.
MBD-LMs achieve a 2x increase in decoding throughput while maintaining high accuracy, revolutionizing the efficiency of diffusion-based text generation.
