0

Video Pre-trained Transformer: A Multimodal Mixture of Pre-trained Experts

Video Pre-trained Transformer uses multiple state-of-the-art encoders with a Flan-T5-11B backbone to generate a universal video representation, incorporating additional modalities like Scene Graphs, for improved performance on downstream tasks.

Year
2023
Venue
arXiv 2023
Authors
4
Hosting
Abstract onlyARXIV-DEFAULT

Cite

Notes

Only stored in your browser.

Attribution

Abstract & full text
arxiv.org/abs/2304.10505ARXIV-DEFAULT
TL;DR
Semantic Scholar
Attribution policy →

Abstract

We present Video Pre-trained Transformer. VPT uses four SOTA encoder models from prior work to convert a video into a sequence of compact embeddings. Our backbone, based on a reference Flan-T5-11B architecture, learns a universal representation of the video that is a non-linear sum of the encoder models. It learns using an autoregressive causal language modeling loss by predicting the words spoken in YouTube videos. Finally, we evaluate on standard downstream benchmarks by training fully connected prediction heads for each task. To the best of our knowledge, this is the first use of multiple frozen SOTA models as encoders in an "embedding -> backbone -> prediction head" design pattern - all others have trained their own joint encoder models. Additionally, we include more modalities than the current SOTA, Merlot Reserve, by adding explicit Scene Graph information. For these two reasons, we believe it could combine the world's best open-source models to achieve SOTA performance. Initial experiments demonstrate the model is learning appropriately, but more experimentation and compute is necessary, and already in progress, to realize our loftier goals. Alongside this work, we build on the YT-20M dataset, reproducing it and adding 25,000 personally selected YouTube videos to its corpus. All code and model checkpoints are open sourced under a standard MIT license.

Authors

4