A Simple Recipe for Contrastively Pre-training Video-First Encoders Beyond 16 Frames

• URL: http://arxiv.org/abs/2312.07395v2
• Date: Mon, 30 Dec 2024 09:51:22 GMT
• Title: A Simple Recipe for Contrastively Pre-training Video-First Encoders Beyond 16 Frames
• Authors: Pinelopi Papalampidi, Skanda Koppula, Shreya Pathak, Justin Chiu, Joe Heyward, Viorica Patraucean, Jiajun Shen, Antoine Miech, Andrew Zisserman, Aida Nematzadeh,
• Abstract summary: We build on the common paradigm of transferring large-scale, image--text models to video via shallow temporal fusion. We expose two limitations to the approach: (1) decreased spatial capabilities, likely due to poor video--language alignment in standard video datasets, and (2) higher memory consumption, bottlenecking the number of frames that can be processed.
• Score: 57.758863967770594
• License:
• Abstract: Understanding long, real-world videos requires modeling of long-range visual dependencies. To this end, we explore video-first architectures, building on the common paradigm of transferring large-scale, image--text models to video via shallow temporal fusion. However, we expose two limitations to the approach: (1) decreased spatial capabilities, likely due to poor video--language alignment in standard video datasets, and (2) higher memory consumption, bottlenecking the number of frames that can be processed. To mitigate the memory bottleneck, we systematically analyze the memory/accuracy trade-off of various efficient methods: factorized attention, parameter-efficient image-to-video adaptation, input masking, and multi-resolution patchification. Surprisingly, simply masking large portions of the video (up to 75%) during contrastive pre-training proves to be one of the most robust ways to scale encoders to videos up to 4.3 minutes at 1 FPS. Our simple approach for training long video-to-text models, which scales to 1B parameters, does not add new architectural complexity and is able to outperform the popular paradigm of using much larger LLMs as an information aggregator over segment-based information on benchmarks with long-range temporal dependencies (YouCook2, EgoSchema).

Related papers

• Video-Panda: Parameter-efficient Alignment for Encoder-free Video-Language Models [26.866184981409607]
  Current video models typically rely on heavyweight image encoders (300M-1.1B parameters) or video encoders (1B-1.4B parameters) Our method introduces a novel Spatio-Temporal Alignment Block (STAB) that directly processes video inputs without requiring pre-trained encoders. Our model achieves comparable or superior performance to encoder-based approaches for open-ended video question answering on standard benchmarks.
  arXiv  Detail & Related papers  (2024-12-24T18:59:56Z)
• Look Every Frame All at Once: Video-Ma$^2$mba for Efficient Long-form Video Understanding with Multi-Axis Gradient Checkpointing [52.050036778325094]
  Video-Ma$2$mba is a novel architecture that incorporates State Space Models (SSMs) within the Mamba-2 framework. Our approach significantly reduces the memory footprint compared to standard gradient checkpointing. By maintaining a detailed capture of temporal dynamics, our model improves the accuracy and relevance of responses in long video understanding tasks.
  arXiv  Detail & Related papers  (2024-11-29T04:12:13Z)
• LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding [65.46303012350207]
  LongVU is an adaptive compression mechanism that reduces the number of video tokens while preserving visual details of long videos. We leverage DINOv2 features to remove redundant frames that exhibit high similarity. We perform spatial token reduction across frames based on their temporal dependencies.
  arXiv  Detail & Related papers  (2024-10-22T21:21:37Z)
• VideoLLaMB: Long-context Video Understanding with Recurrent Memory Bridges [42.555895949250704]
  VideoLLaMB is a novel framework that utilizes temporal memory tokens within bridge layers to allow for the encoding of entire video sequences. SceneTilling algorithm segments videos into independent semantic units to preserve semantic integrity. In terms of efficiency, VideoLLaMB, trained on 16 frames, supports up to 320 frames on a single Nvidia A100 GPU.
  arXiv  Detail & Related papers  (2024-09-02T08:52:58Z)
• Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization [52.63845811751936]
  Video pre-training is challenging due to the modeling of its dynamics video. In this paper, we address such limitations in video pre-training with an efficient video decomposition. Our framework is both capable of comprehending and generating image and video content, as demonstrated by its performance across 13 multimodal benchmarks.
  arXiv  Detail & Related papers  (2024-02-05T16:30:49Z)
• Memory Efficient Temporal & Visual Graph Model for Unsupervised Video Domain Adaptation [50.158454960223274]
  Existing video domain adaption (DA) methods need to store all temporal combinations of video frames or pair the source and target videos. We propose a memory-efficient graph-based video DA approach.
  arXiv  Detail & Related papers  (2022-08-13T02:56:10Z)
• 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)

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.