HSE progressively distills procedural knowledge from coding trajectories through hierarchical evidence aggregation, which evolves the skill library from target-benchmark trajectories without verifier feedback, relying solely on the learned evaluator for candidate assessment and achieving additional gains through benchm...
Jun Gao, Yun Peng, Chang-Hai Zhou et al.· 0 citations
This work introduces MJev-Spatial, a System 1 multimodal decision model that tackles all three types of spatial understanding tasks within a unified classification framework, and achieves the highest overall score among eight tested models and outperforms its unmodified autoregressive base model on all twelve benchmark...
Luo-Zheng Qin, Yu-Han Zhu, Chang-Hai Zhou et al.· 1 citation
OneStreamer is introduced, which jointly learns query-independent evidence recording and task response through a shared proactive generation process, and the 4B model achieves the best results among the compared methods across all eight evaluated streaming video understanding benchmarks.
VideoJev, a family of models that cast video understanding and interaction as structured decision tasks, is introduced, a family of models that cast video understanding and interaction as structured decision tasks that combine the semantic flexibility of MLLMs with efficient decisions that agents can use directly.
Yu-Han Zhu, Xiang-Yun Zeng, Qing-Yi Si et al.· 1 citation
Across six real-world scenarios, human raters prefer JoyAI-VL-Interaction over the in-app video-call assistants of Doubao and Gemini by a wide margin, and the first open, vision-driven interaction model released together with its training recipe, data, and complete deployable system.
Ding-Yu Yao, Jun Zhou, Chenxu Yang et al.· arXiv.org· 10 citations
VIABench is introduced, a comprehensive video benchmark specifically designed to evaluate MLLMs in Visually Impaired Assistance scenarios using first-person videos recorded or shared by VIIs themselves, and proposes a rigorous benchmarking pipeline that supports both online (real-time) and offline settings.
Yunfeng Liu, Yuandong Yang, Jiarui Han et al.· arXiv.org· 0 citations
This work introduces VideoChat3, a fully open, efficient, and generalist video-centric MLLM, which surpasses prior open-source models with equal or larger parameter counts with only 4B parameters and higher efficiency.
Across seven benchmarks, TimeLens2-2B outperforms all size-matched baselines on every benchmark, while the 4B and 8B variants achieve state-of-the-art performance, surpassing open-source models with up to 397B parameters.