Yuan Gao

Long-term identity preservation in multi-object tracking is far more challenging than previously understood, with all tested methods exhibiting substantial fragmentation in trajectories.

World models may be fundamentally misinterpreting future states, imagining them kinematically rather than dynamically, leading to significant long-horizon failures.

TRCGL-Net achieves a remarkable tail-class mAP of 0.4904, setting a new benchmark for rare disease recognition in chest X-ray classification.

ClinRAG-GRAPH achieves impressive pCR prediction accuracy while ensuring interpretability and robustness against imaging biases across multiple centers.

Achieving state-of-the-art kanji reading accuracy, Sarashina2.2-TTS sets a new standard for Japanese speech synthesis while ensuring cross-lingual robustness.

Clarifying memories can significantly boost the factual accuracy and personalization of conversational agents, while irrelevant memories lead to degraded responses.

WVM outperforms existing models by accurately assessing task progressions and improving robotic manipulation from both expert and suboptimal data.

A 4B-parameter model outperforms Gemini-3-Pro in autonomous driving by incorporating physics-informed constraints and style-aware training, suggesting specialized models can surpass larger, general-purpose models in domain-specific tasks.

Sports expose surprising limitations in VLMs' spatial reasoning, as current models struggle to generalize from existing benchmarks despite fine-tuning gains on a new, large-scale dataset.