https://arxiv.org/abs/2509.03951v1
[2509.03951v1] ANTS: Shaping the Adaptive Negative Textual Space by MLLM for OOD Detection
Abstract page for arXiv paper 2509.03951v1: ANTS: Shaping the Adaptive Negative Textual Space by MLLM for OOD Detection
https://proceedings.nips.cc/paper_files/paper/2025/hash/19e4ea30dded58259665db375885e412-Abstract-Datasets_and_Benchmarks_Track.html
RTV-Bench: Benchmarking MLLM Continuous Perception, Understanding and Reasoning through Real-Time...
https://proceedings.iclr.cc/paper_files/paper/2025/hash/24079b91da7257cb78805262996152b8-Abstract-Conference.html
MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation
mllmseedynamiccorrectiondecoding
https://arxiv.org/abs/2506.14766v2
[2506.14766v2] ASCD: Attention-Steerable Contrastive Decoding for Reducing Hallucination in MLLM
Abstract page for arXiv paper 2506.14766v2: ASCD: Attention-Steerable Contrastive Decoding for Reducing Hallucination in MLLM
https://aiiiii.com.cn/sites/8424.html
MG-LLaVA: 增强视觉处理能力的机器学习语言模型(MLLM) | AI设计导航
MG-LLaVA是一个增强视觉处理能力的机器学习语言模型（MLLM），通过整合多粒度视觉流程，包括低分辨率、高分辨率和以对象为中心的特征。提出了一个额外的高分辨率视觉编码器来捕捉细节，并通过Conv-Gate融合网络与基础视觉特征融合。
mgllavamllm