Towards Fast and Effective Long Video Understanding of Multimodal Large Language Models via Adaptive Quasi-Gaussian Sampling

• URL: http://arxiv.org/abs/2606.24187v2
• Date: Wed, 24 Jun 2026 04:06:43 GMT
• Title: Towards Fast and Effective Long Video Understanding of Multimodal Large Language Models via Adaptive Quasi-Gaussian Sampling
• Abstract summary: Video frame selection is a problem of Quasi-Gaussian Sampling.<n>We propose an adaptive and training-free approach termed AdaQ.
• Score: 71.62893745623526
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Long video understanding remains a daunting challenge for Multimodal Large Language Models (MLLMs) due to the excessive computation and memory footprint. Thus, keyframe selection is often adopted to mitigate this shortcoming, which however still suffers from low flexibility and high noise due to its hard sampling principle. In this paper, we define video frame selection as a problem of Quasi-Gaussian Sampling, and propose an adaptive and training-free approach termed AdaQ. Inspired by the 3-$σ$ rule of Gaussian distribution, the objective of AdaQ is to achieve the optimal 3-$σ$ interval for different examples, i.e., a smaller 3-$σ$ interval for the local query and a larger one for the global query, thereby facilitating robust and adaptive frame sampling. To validate AdaQ, we apply it to four MLLMs with three embedding models. The extensive experimental results not only show its obvious performance gains over the default MLLMs and the SOTA keyframe selection methods, e.g., helping Qwen3-VL-8B outperform GPT4o by 15.8% on average by using only 64 frames, but also confirm its superior robustness and high efficiency for long-video understanding, e.g., only 1 hyper-parameter needs to be set.

Related papers

• MarKey: Marginal Utility Guided Greedy Keyframe Selection for Long Video Understanding [17.82627304498428]
  Long-video understanding remains challenging for large language models.<n>Recent training-free selection methods have enabled more efficient inference.<n>We propose MarKey, a framework that formulates multimodal selection as subset-aware greedy optimization.
  arXiv  Detail & Related papers  (2026-09-14T11:39:20Z)
• Efficient Frame Selection for Long Videos at Test Time with Attention-Based MLLM Selectors [69.75895255395702]
  Cross-modal attention at validation-selected extraction layers in MLLMs already provides query-relevant frame evidence without requiring autoregressive generation.<n>A lightweight MLLM selector, even with only 2B parameters, can extract frame-level evidence by converting selected-layer attention into relevance scores through query-conditioned aggregation.<n>Under a 32-frame budget, our selector improves over uniform sampling by up to 6.4 points on Video-MME and outperforms prior training-based selectors under matched frame budgets.
  arXiv  Detail & Related papers  (2026-07-17T07:04:31Z)
• Gaussian Mixture Modeling for Event-Aware Visual Allocation in Long Video Understanding [55.04091863981529]
  Large Vision-Language Models (LVLMs) face significant challenges in long video understanding due to the excessive computational cost and information loss associated with uniform sampling.<n>We propose GMM-EVA, which leverages Gaussian Mixture Models to model event-level structure from discrete frame-wise.<n>A differentiated allocation strategy is then applied to preserve one primary high-resolution per event for high-fidelity detail, while utilizing lower-resolution secondary observations to maintain temporal context and optimize token budgets.
  arXiv  Detail & Related papers  (2026-07-14T09:27:25Z)
• GridProbe: Posterior-Probing for Adaptive Test-Time Compute in Long-Video VLMs [3.9266376632068485]
  GridProbe is an efficient training-free posterior-probing inference paradigm.<n>It scores evidence in answer space using a frozen VLM's own reasoning.<n>It selects question-relevant frames adaptively, resulting in sub-quadratic attention cost with little to no accuracy loss.
  arXiv  Detail & Related papers  (2026-05-11T15:57:46Z)
• One Token per Highly Selective Frame: Towards Extreme Compression for Long Video Understanding [51.08792182064565]
  Long video understanding is inherently challenging for vision-language models (VLMs) because of the extensive number of frames.<n>With each video frame typically expanding into tens or hundreds of tokens, the limited context length of large language models (LLMs) forces the VLMs to perceive the frames sparsely and lose temporal information.<n>We propose an extreme compression model for long video understanding, named XComp, achieving a significantly larger compression ratio and enabling denser frame sampling.
  arXiv  Detail & Related papers  (2026-04-15T17:59:52Z)
• Query-Conditioned Evidential Keyframe Sampling for MLLM-Based Long-Form Video Understanding [21.306038832082553]
  Multimodal Large Language Models (MLLMs) have shown strong performance on video question answering.<n>Their application to long-form videos is constrained by limited context length and computational cost.<n>We propose an evidence-driven sampling framework grounded in information theory.
  arXiv  Detail & Related papers  (2026-04-01T15:02:50Z)
• Separating Oblivious and Adaptive Models of Variable Selection [13.61388474201292]
  We show that the optimal $ell_infty$ error is attainable in near-linear time with $gtrsim k2$ samples.<n>We conclude with a preliminary examination of a emphpartially-adaptive $ model.
  arXiv  Detail & Related papers  (2026-02-18T16:10:35Z)
• Scoring, Reasoning, and Selecting the Best! Ensembling Large Language Models via a Peer-Review Process [58.265053900416895]
  LLM-PeerReview is built on a novel, peer-review-inspired framework.<n>It operates in three stages: For scoring, we use the emerging LLM-as-a-Judge technique.<n>For reasoning, we can apply a graphical model-based truth inference algorithm.<n>Finally, the highest-scoring response is selected as the best ensemble output.
  arXiv  Detail & Related papers  (2025-12-29T05:25:49Z)
• Test-Time Temporal Sampling for Efficient MLLM Video Understanding [26.144261085897863]
  Test-Time Temporal Sampling (T3S) is a training-free, plug-and-play inference wrapper that enables MLLMs to process long videos both efficiently and effectively.<n>Our approach operates entirely at inference time, requires no model modifications or fine-tuning, and is compatible with a wide range of pretrained MLLMs.
  arXiv  Detail & Related papers  (2025-11-22T06:59:21Z)
• LOVE-R1: Advancing Long Video Understanding with an Adaptive Zoom-in Mechanism via Multi-Step Reasoning [73.90466023069125]
  We propose LOVE-R1, a model that can adaptively zoom in on a video clip.<n>The model is first provided with densely sampled frames but in a small resolution.<n>If some spatial details are needed, the model can zoom in on a clip of interest with a large frame resolution.
  arXiv  Detail & Related papers  (2025-09-29T13:43:55Z)
• PEVLM: Parallel Encoding for Vision-Language Models [4.777805570120456]
  We introduce textbfPEVLM, a fine-tuning-free parallel encoding method designed to enhance the prefilling efficiency of Vision-Language Models.<n>PEVLM partitions the input video into context blocks with a shared sink block, while preserving sequential position embeddings to align the attention weight distribution with that of Full-Attention.<n>Experiments demonstrate that PEVLM consistently outperforms existing parallel encoding approaches, achieving up to textbf7.47x speedup in attention computation and reducing end-to-end latency by textbf40%.
  arXiv  Detail & Related papers  (2025-06-24T14:14:52Z)
• Metropolis-Hastings Sampling for 3D Gaussian Reconstruction [31.840492077537018]
  We propose an adaptive sampling framework for 3D Gaussian Splatting (3DGS)<n>Our framework overcomes limitations by reformulating densification and pruning as a probabilistic sampling process.<n>Our approach achieves faster convergence while matching or modestly surpassing the view-synthesis quality of state-of-the-art models.
  arXiv  Detail & Related papers  (2025-06-15T19:12:37Z)
• Adaptive Keyframe Sampling for Long Video Understanding [75.7837692594814]
  This paper presents a simple yet effective algorithm named Adaptive Keyframe Sampling (AKS)<n>It inserts a plug-and-play module known as Adaptive Keyframe Sampling (AKS) which aims to maximize the useful information with a fixed number of video tokens.<n>Experiments on two long video understanding benchmarks validate that AKS improves video QA accuracy upon selecting informative encounters.
  arXiv  Detail & Related papers  (2025-02-28T17:46:29Z)
• Sign Operator for Coping with Heavy-Tailed Noise in Non-Convex Optimization: High Probability Bounds Under $(L_0, L_1)$-Smoothness [74.18546828528298]
  We show that SignSGD with Majority Voting can robustly work on the whole range of complexity with $kappakappakappakappa-1right, kappakappakappa-1right, kappakappakappa-1right, kappakappakappa-1right, kappakappakappa-1right, kappakappakappa-1right, kappakappakappa-1right, kappa
  arXiv  Detail & Related papers  (2025-02-11T19:54:11Z)
• AMOM: Adaptive Masking over Masking for Conditional Masked Language Model [81.55294354206923]
  A conditional masked language model (CMLM) is one of the most versatile frameworks. We introduce a simple yet effective adaptive masking over masking strategy to enhance the refinement capability of the decoder. Our proposed model yields state-of-the-art performance on neural machine translation.
  arXiv  Detail & Related papers  (2023-03-13T20:34:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.