TensorX
返回文献探索

Paper · arXiv 2402.13217

VideoPrism: A Foundational Visual Encoder for Video Understanding

Long Zhao, Nitesh B. Gundavarapu, Liangzhe Yuan, Hao Zhou, Shen Yan, Jennifer J. Sun, Luke Friedman, Rui Qian, Tobias Weyand, Yue Zhao, Rachel Hornung, Florian Schroff, Ming-Hsuan Yang, David A. Ross, Huisheng Wang, Hartwig Adam, Mikhail Sirotenko, Ting Liu, Boqing Gong

41 upvotesFebruary 20, 2024arXiv 预印本
AI 摘要

VideoPrism, a pretrained video encoder, achieves top performance across various video understanding tasks by utilizing global-local distillation and token shuffling of semantic video embeddings enhanced with associated text.

VideoPrismmasked autoencodingglobal-local distillationsemantic video embeddingstoken shuffling schemevideo understanding tasksweb video question answeringCV for sciencevideo understanding benchmarks

Abstract

We introduce VideoPrism, a general-purpose video encoder that tackles diverse video understanding tasks with a single frozen model. We pretrain VideoPrism on a heterogeneous corpus containing 36M high-quality video-caption pairs and 582M video clips with noisy parallel text (e.g., ASR transcripts). The pretraining approach improves upon masked autoencoding by global-local distillation of semantic video embeddings and a token shuffling scheme, enabling VideoPrism to focus primarily on the video modality while leveraging the invaluable text associated with videos. We extensively test VideoPrism on four broad groups of video understanding tasks, from web video question answering to CV for science, achieving state-of-the-art performance on 30 out of 33 video understanding benchmarks.

北京市昌平区探索星信息技术及软件开发工作室

京ICP备2026059466号
VideoPrism: A Foundational Visual Encoder for Video Understanding | TensorX