Search papers, labs, and topics across Lattice.
Department of Computer Science and Engineering, IIT Delhi
3
0
7
Pretraining ViTs with DINOv2 and focusing attention on regions of interest dramatically improves breast cancer detection from mammograms, outperforming standard approaches.
LLMs struggle to identify relevant legal issues with high precision, but a neuro-symbolic approach using sparse linear models over LLM-generated analytical factors can boost performance by 30-40%.
Transformer-based SLAM can achieve state-of-the-art accuracy and efficiency by integrating a spatially corrective back-end that restores high-cadence local bundle adjustment.
