Yue Zhang

PP-OCRv6 outperforms billion-scale VLMs on OCR tasks with a fraction of the parameters, achieving state-of-the-art accuracy and speed.

AutoMine outperforms existing methods in scenario mining, achieving a HOTA-Temporal score of 36.38 in a competitive setting.

Cross-layer attention aggregation in LASA boosts semantic segmentation accuracy by over 15% in challenging sketch datasets, revealing the untapped potential of multi-layer features.

EviProp achieves superior evidence-page retrieval by leveraging a novel graph-based approach that captures complex document structures, outperforming traditional methods.

AutoPilot slashes consensus latency by nearly 50% through intelligent, real-time parameter tuning in Byzantine Fault Tolerant protocols.

UNIVID cuts violation leakage by 42.7% while consolidating over 1,000 classifiers into a single, interpretable model for video moderation.

Traditional aesthetic assessment models miss the mark by relying on static scores, while RED-Aes reveals the dynamic, comparative nature of human aesthetic perception for superior generalization.

SlidingServe achieves a 30% increase in service capacity while slashing SLO violations by up to 53%, revolutionizing LLM inference scheduling.

Targeted optimization in underperforming regions boosts document parsing accuracy to a record 96.33%, setting a new benchmark in the field.

VLMs confidently hallucinate answers to spatial reasoning questions even when visual evidence is occluded or misleading, achieving near-random performance in identifying viewpoints that could resolve the ambiguity.

Forget backprop and memory lookups: FAAST lets you adapt models at test time with a single forward pass, matching fine-tuning accuracy with massive speed and memory gains.

LLM-based peer review systems can be made significantly more robust against adversarial manipulation via a co-evolutionary GAN approach that anticipates novel attacks.

Current identity management systems fail for AI agents, but AgentDID offers a scalable, decentralized solution that lets agents manage their own identities and prove their state at interaction time.