This paper introduce the LLM-PySC2 environment, an LLM decision-making environment with complete pysc2 action space and multi-agent system.

We propose MultiFCL, a novel FCL framework that fine-tunes PTMs to adapt to FCL while preserving their strong generalization capabilities.

TempSamp-R1 is a reinforcement fine-tuning framework that integrates off-policy supervision, soft advantage shaping, and hybrid Chain-of-Thought training to enhance the temporal grounding capabilities of MLLMs.

This paper proposes to add scene-centric forecasting as an effective control to enhance occupancy world models.

We present an emotion-centric video foundation model trained with fine-grained captions and rationales via affective-tree reasoning guidance, achieving high-level emotional intelligence for video understanding.

This paper explores using causal visual prompts to address spurious correlations in single-domain generalization

This paper presents SeniorTalk, a Chinese conversation dataset focused on speech from people aged 75 and older, supporting various speech tasks.

We propose Cradle2Cane, a novel two-pass diffusion-based face aging method that balances age accuracy and identity preservation, achieving state-of-the-art performance on the CelebA test dataset with improved efficiency.

We propose a new perspective to analyse the memory bank rewriting, and introduce a new inter-class constraint.

Exploring and Mitigating hallucination in LMMs towards scene text spotting and understanding