SciVid: Cross-Domain Evaluation of Video Models in Scientific Applications
This work addresses the need for cross-domain evaluation of video models in scientific applications, providing a benchmark to assess generalizability, though it is incremental in adapting existing models to new tasks.
The paper tackles the problem of evaluating whether video foundation models (ViFMs) can transfer knowledge across diverse scientific disciplines, introducing the SciVid benchmark with five tasks in medical, animal behavior, and weather domains. It shows that adapted ViFMs achieve state-of-the-art results in several applications, with specific performance gains, while also revealing limitations in existing models.
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.