5 papers
ReVEAL: Eliciting Role-Aware Latent Visual Evidence from MLLMs for Reasoning Video Object Segmentation
Zhongyuan Bao, Cunjie He, Lejun Zhang +2
Reasoning video object segmentation (ReVOS) requires segmenting the target specified by an implicit query throughout a video. Existing MLLM-based methods typically require costly c…
Open-o3-Video-v2: Seek, Ground, and Reason with Verifiable Spatio-Temporal Evidence
Jiahao Meng, Zhongyuan Bao, Yue Tan +7
Reliable video reasoning requires models to produce correct answers and expose supporting visual evidence. Open-o3-Video enables verifiable reasoning by embedding timestamps and ob…
Lotus: Efficient LLM Training by Randomized Low-Rank Gradient Projection with Adaptive Subspace Switching
Tianhao Miao, Zhongyuan Bao, Lejun Zhang
Training efficiency in large-scale models is typically assessed through memory consumption, training time, and model performance. Current methods often exhibit trade-offs among the…
CrossCheck-Bench: Diagnosing Compositional Failures in Multimodal Conflict Resolution
Baoliang Tian, Yuxuan Si, Jilong Wang +13
Multimodal Large Language Models are primarily trained and evaluated on aligned image-text pairs, which leaves their ability to detect and resolve real-world inconsistencies largel…
TennisTV: Do Multimodal Large Language Models Understand Tennis Rallies?
Zhongyuan Bao, Lejun Zhang
Multimodal large language models (MLLMs) excel at general video understanding but struggle with fast, high-frequency sports like tennis, where rally clips are short yet information…