Video-MME-v2
PulseAugur coverage of Video-MME-v2 — every cluster mentioning Video-MME-v2 across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
FrameRouter improves long-video understanding with selective frame budgeting
Researchers have developed FrameRouter, a novel framework designed to improve long-video understanding in question-answering tasks. This system addresses the inefficiency of uniform frame budgets by dynamically allocati…
-
SleuthVLM research shows agentic reasoning surpasses audio in video understanding
A new research paper introduces SleuthVLM, an agentic framework for video understanding that bypasses direct audio processing. Instead, SleuthVLM transcribes speech to text and uses a vision-language model to analyze th…
-
New benchmarks and methods enhance audio-visual reasoning in LLMs · 5 sources tracked
Researchers have developed new methods and benchmarks to improve audio-visual joint reasoning in large language models. OmniReasoningBench and OmniSeek are introduced as frameworks and datasets designed to evaluate and …
-
Kwai releases Keye-VL-2.0 for long-video understanding
Kwai has released Keye-VL-2.0-30B-A3B, an open-source multimodal foundation model designed for long-video understanding and agentic intelligence. This model utilizes DeepSeek Sparse Attention to process up to 256K conte…
-
GridProbe cuts VLM compute cost for long videos
Researchers have developed GridProbe, a novel method to improve the efficiency of long-video Visual Language Models (VLMs). This technique adaptively selects relevant frames during inference, reducing the computational …