SciVid: Cross-Domain Evaluation of Video Models in Scientific Applications

TL;DR

This paper introduces SciVid, a benchmark for evaluating video foundation models across scientific domains, demonstrating their transferability and identifying limitations for future development.

Contribution

It presents SciVid, a new benchmark with five scientific video tasks, and evaluates six ViFMs, showing their potential and limitations in scientific applications.

Findings

State-of-the-art results achieved in several scientific tasks using ViFMs

Transfer learning from large-scale data is effective across disciplines

Limitations of current ViFMs highlight need for more generalizable models

Abstract

In recent years, there has been a proliferation of spatiotemporal foundation models in different scientific disciplines. While promising, these models are often domain-specific and are only assessed within the particular applications for which they are designed. Given that many tasks can be represented as video modeling problems, video foundation models (ViFMs) hold considerable promise as general-purpose domain-agnostic approaches. However, it is not known whether the knowledge acquired on large-scale but potentially out-of-domain data can be effectively transferred across diverse scientific disciplines, and if a single, pretrained ViFM can be competitive with domain-specific baselines. To address this, we introduce SciVid, a comprehensive benchmark comprising five *Sci*entific *Vid*eo tasks, across medical computer vision, animal behavior, and weather forecasting. We adapt six leading ViFMs to SciVid using simple trainable readout modules, establishing strong baselines and demonstrating the potential for effective transfer learning. Specifically, we show that state-of-the-art results can be obtained in several applications by leveraging the general-purpose representations from ViFM backbones. Furthermore, our results reveal the limitations of existing ViFMs, and highlight opportunities for the development of generalizable models for high-impact scientific applications. We release our code at https://github.com/google-deepmind/scivid to facilitate further research in the development of ViFMs.