🏢 Peking University

Code-as-Monitor (CaM) uses vision-language models and constraint-aware visual programming to achieve both reactive and proactive robotic failure detection in real-time, improving success rates and red…

Imperfect OCR hinders Retrieval-Augmented Generation (RAG). OHRBench, a new benchmark, reveals this cascading impact, showing current OCR solutions insufficient for high-quality RAG knowledge bases. …

Open-Sora Plan introduces an open-source large video generation model capable of producing high-resolution videos with long durations, based on various user inputs.

VideoLLM’s interaction format is revolutionized by the novel Video-Text Duet, enabling real-time, time-sensitive video comprehension with significantly improved performance.

WF-VAE boosts video VAE performance with wavelet-driven energy flow and causal caching, enabling 2x higher throughput and 4x lower memory usage in latent video diffusion models.

ConsisID achieves high-quality, identity-preserving text-to-video generation using a tuning-free diffusion transformer model that leverages frequency decomposition for effective identity control.

LLaVA-01: A novel visual language model achieves superior reasoning performance through structured, multi-stage processing and efficient inference-time scaling, surpassing even larger, closed-source m…

LLMs can now self-improve long-context reasoning via SEALONG, a novel method leveraging multiple model outputs and minimum Bayes risk scoring to enable effective supervised fine-tuning or preference o…

GaussianAnything: Interactive point cloud latent diffusion enables high-quality, editable 3D models from images or text, overcoming existing 3D generation limitations.

KMM: Key Frame Mask Mamba generates extended, diverse human motion from text prompts by innovatively masking key frames in the Mamba architecture and using contrastive learning for improved text-motio…

Training-free Regional Prompting for FLUX boosts compositional text-to-image generation by cleverly manipulating attention mechanisms, achieving fine-grained control without retraining.

DreamPolish: A new text-to-3D model generates highly detailed 3D objects with polished surfaces and realistic textures using progressive geometry refinement and a novel domain score distillation tech…

HelloMeme enhances text-to-image models by integrating spatial knitting attentions, enabling high-fidelity meme video generation while preserving model generalization.