LookWhen? Fast Video Recognition by Learning When, Where, and What to Compute

• URL: http://arxiv.org/abs/2605.06809v1
• Date: Thu, 07 May 2026 18:08:31 GMT
• Title: LookWhen? Fast Video Recognition by Learning When, Where, and What to Compute
• Abstract summary: We introduce LookWhen, a selector-extractor framework that factorizes video recognition into learning when, where, and what to compute.<n>Our shallow selector gets a scaled-down video and quickly scores all tokens across space-time, while our deep extractor gets the top-K selected tokens to approximate full-video representations.<n>LookWhen is more efficient still at 6.7x faster than InternVideo2-B at equal accuracy.
• Score: 32.29218279577984
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Transformers dominate video recognition. They split videos into tokens, and processing them has expensive superlinear computational cost. Yet videos are filled with redundancy, so we can question the need for this expense. We introduce LookWhen, a selector-extractor framework that factorizes video recognition into learning when, where, and what to compute. Our shallow selector gets a scaled-down video and quickly scores all tokens across space-time, while our deep extractor gets the top-K selected tokens to approximate full-video representations without actually processing all the tokens. A key challenge is defining effective supervision for selection and extraction. For selection pre-training, we introduce a score on representations that ranks tokens by uniqueness using a simple nearest-neighbor distance. For extraction pre-training, we distill both a video teacher and an image teacher, for which we normalize its frame-wise representations to learn what changes within videos. Through these strategies, our selector-extractor learns general and efficient representations for feature extraction or fine-tuning to a task. Through experiments on Kinetics-400, SSv2, Epic-Kitchens, Diving48, Jester, and Charades, we show that LookWhen achieves a better accuracy-computation trade-off than efficient models and upgraded baselines of similar size. LookWhen Pareto-dominates in accuracy-FLOPs on 9 of 12 cases (6 tasks x 2 settings) and roughly matches on 3. In accuracy-throughput, measuring time in practice, LookWhen is more efficient still at 6.7x faster than InternVideo2-B at equal accuracy.

Related papers

• LookWhere? Efficient Visual Recognition by Learning Where to Look and What to See from Self-Supervision [10.461453853510964]
  Vision transformers are ever larger, more accurate, and more expensive to compute.<n>We turn to adaptive computation to cope with this cost by learning to predict where to compute.<n>Our LookWhere method divides a low-resolution selector and a high-resolution extractor without ever processing the full high-resolution input.
  arXiv  Detail & Related papers  (2025-05-23T15:56:35Z)
• How can objects help action recognition? [74.29564964727813]
  We investigate how we can use knowledge of objects to design better video models. First, we propose an object-guided token sampling strategy that enables us to retain a small fraction of the input tokens. Second, we propose an object-aware attention module that enriches our feature representation with object information.
  arXiv  Detail & Related papers  (2023-06-20T17:56:16Z)
• Efficient Video Action Detection with Token Dropout and Context Refinement [67.10895416008911]
  We propose an end-to-end framework for efficient video action detection (ViTs) In a video clip, we maintain tokens from its perspective while preserving tokens relevant to actor motions from other frames. Second, we refine scene context by leveraging remaining tokens for better recognizing actor identities.
  arXiv  Detail & Related papers  (2023-04-17T17:21:21Z)
• EVEREST: Efficient Masked Video Autoencoder by Removing Redundant Spatiotemporal Tokens [57.354304637367555]
  We present EVEREST, a surprisingly efficient MVA approach for video representation learning. It finds tokens containing rich motion features and discards uninformative ones during both pre-training and fine-tuning. Our method significantly reduces the computation and memory requirements of MVA.
  arXiv  Detail & Related papers  (2022-11-19T09:57:01Z)
• Efficient Video Transformers with Spatial-Temporal Token Selection [68.27784654734396]
  We present STTS, a token selection framework that dynamically selects a few informative tokens in both temporal and spatial dimensions conditioned on input video samples. Our framework achieves similar results while requiring 20% less computation.
  arXiv  Detail & Related papers  (2021-11-23T00:35:58Z)
• ASCNet: Self-supervised Video Representation Learning with Appearance-Speed Consistency [62.38914747727636]
  We study self-supervised video representation learning, which is a challenging task due to 1) a lack of labels for explicit supervision and 2) unstructured and noisy visual information. Existing methods mainly use contrastive loss with video clips as the instances and learn visual representation by discriminating instances from each other. In this paper, we observe that the consistency between positive samples is the key to learn robust video representations.
  arXiv  Detail & Related papers  (2021-06-04T08:44:50Z)
• Beyond Short Clips: End-to-End Video-Level Learning with Collaborative Memories [56.91664227337115]
  We introduce a collaborative memory mechanism that encodes information across multiple sampled clips of a video at each training iteration. This enables the learning of long-range dependencies beyond a single clip. Our proposed framework is end-to-end trainable and significantly improves the accuracy of video classification at a negligible computational overhead.
  arXiv  Detail & Related papers  (2021-04-02T18:59:09Z)
• Temporal Stochastic Softmax for 3D CNNs: An Application in Facial Expression Recognition [11.517316695930596]
  We present a strategy for efficient video-based training of 3D CNNs. It relies on softmax temporal pooling and a weighted sampling mechanism to select the most relevant training clips.
  arXiv  Detail & Related papers  (2020-11-10T16:40:00Z)
• RSPNet: Relative Speed Perception for Unsupervised Video Representation Learning [100.76672109782815]
  We study unsupervised video representation learning that seeks to learn both motion and appearance features from unlabeled video only. It is difficult to construct a suitable self-supervised task to well model both motion and appearance features. We propose a new way to perceive the playback speed and exploit the relative speed between two video clips as labels.
  arXiv  Detail & Related papers  (2020-10-27T16:42:50Z)

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.