Abstract
Frame selection is an important component of long-video question answering (VQA) with Multimodal Large Language Models (MLLMs). Existing frame-selection methods improve over simple top-$k$ embedding retrieval and uniform sampling, but are typically applied under a fixed global selection budget. We introduce \textbf{MetaSampling}, a training-free, plug-and-play sampling strategy that can be applied on top of existing frame selectors. MetaSampling improves downstream VQA efficiency by dynamically reducing the number of frames passed to the MLLM while preserving, and in some cases improving, answer accuracy. We evaluate MetaSampling across 36 paired frame-selector--MLLM-backbone--VQA-benchmark configurations. MetaSampling reduces the number of selected frames in all 36 configurations and improves accuracy in 25 of them, yielding an average frame reduction of $8.9\%$ while slightly improving accuracy overall.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Dahal, A., & Banerjee, B. (2026). MetaSampling: Making Frame Samplers Efficient for Long-Video Question Answering. https://omanscience.com/en/articles/metasampling-making-frame-samplers-efficient-for-long-video-question-answering
MLA 9
Dahal, Ashim, and Bikramjit Banerjee. "MetaSampling: Making Frame Samplers Efficient for Long-Video Question Answering." https://omanscience.com/en/articles/metasampling-making-frame-samplers-efficient-for-long-video-question-answering.
Chicago (author–date)
Dahal, Ashim, and Bikramjit Banerjee. 2026. "MetaSampling: Making Frame Samplers Efficient for Long-Video Question Answering." https://omanscience.com/en/articles/metasampling-making-frame-samplers-efficient-for-long-video-question-answering.
Harvard
Dahal, A. and Banerjee, B. (2026) 'MetaSampling: Making Frame Samplers Efficient for Long-Video Question Answering', Available at: https://omanscience.com/en/articles/metasampling-making-frame-samplers-efficient-for-long-video-question-answering.
Vancouver
Dahal A, Banerjee B. MetaSampling: Making Frame Samplers Efficient for Long-Video Question Answering. https://omanscience.com/en/articles/metasampling-making-frame-samplers-efficient-for-long-video-question-answering
IEEE
A. Dahal, and B. Banerjee, "MetaSampling: Making Frame Samplers Efficient for Long-Video Question Answering," https://omanscience.com/en/articles/metasampling-making-frame-samplers-efficient-for-long-video-question-answering.