Litcius/Paper detail

MAGVIT: Masked Generative Video Transformer

Lijun Yu, Yong Cheng, Kihyuk Sohn, José Lezama, Han Zhang, Hui‐Wen Chang, Alexander G. Hauptmann, Ming–Hsuan Yang, Hao Yuan, Irfan Essa, Lu Jiang

2023106 citationsDOI

Abstract

We introduce the MAsked Generative VIdeo Transformer, MAGVIT, to tackle various video synthesis tasks with a single model. We introduce a 3D tokenizer to quantize a video into spatial-temporal visual tokens and propose an embedding method for masked video token modeling to facilitate multi-task learning. We conduct extensive experiments to demonstrate the quality, efficiency, and flexibility of MAGVIT. Our experiments show that (i) MAGVIT performs favorably against state-of-the-art approaches and establishes the best-published FVD on three video generation benchmarks, including the challenging Kinetics-600. (ii) MAGVIT outperforms existing methods in inference time by two orders of magnitude against diffusion models and by 60x against autoregressive models. (iii) A single MAGVIT model supports ten diverse generation tasks and generalizes across videos from different visual domains. The source code and trained models will be released to the public at https://magvit.cs.cmu.edu.

Topics & Concepts

Computer scienceSecurity tokenTransformerEmbeddingInferenceGenerative modelAutoregressive modelArtificial intelligenceGenerative grammarFlexibility (engineering)Source codeSpeech recognitionMachine learningPattern recognition (psychology)Programming languageComputer securityMathematicsPhysicsStatisticsEconometricsVoltageEconomicsQuantum mechanicsGenerative Adversarial Networks and Image SynthesisVideo Analysis and SummarizationAdvanced Vision and Imaging
MAGVIT: Masked Generative Video Transformer | Litcius