Yadan Luo

Spectral analysis reveals that tokens with dynamic cross-layer evolution are crucial for preserving semantic integrity, leading to more efficient MLLMs.

Long-horizon planning in autonomous driving just got a major upgrade, with GraphWorld cutting collision rates and boosting performance in complex environments.

GraphBEV++ outperforms five existing baselines in addressing critical misalignment issues, significantly boosting performance in both perception and planning tasks for autonomous vehicles.

Text-based prototypes in vision-language models are fundamentally misaligned with visual data for out-of-distribution detection, but this can be overcome with a novel online pseudo-supervised approach.

Decomposing holistic visual cues into subtle, spatially-associated discrepancies allows for state-of-the-art ultra-fine-grained classification even with limited training data.

Soybean leaves have intricate vein structures that unlock state-of-the-art ultra-fine-grained visual categorization, even with limited data.

Achieve SOTA wheat disease segmentation with limited data by cleverly combining the semantic power of DINOv2 with the geometric precision of SAM.