Beyond Instructional Videos: Probing for More Diverse Visual-Textual Grounding on YouTube
This work addresses the problem of limited video understanding diversity for AI researchers, though it is incremental as it extends existing methods to new data.
The researchers investigated whether visual-textual grounding models, previously limited to instructional videos, could be trained on more diverse video corpora like YouTube8M, finding that such grounding is possible across new categories and improves generalization to both non-instructional and instructional domains.
Pretraining from unlabelled web videos has quickly become the de-facto means of achieving high performance on many video understanding tasks. Features are learned via prediction of grounded relationships between visual content and automatic speech recognition (ASR) tokens. However, prior pretraining work has been limited to only instructional videos; a priori, we expect this domain to be relatively "easy:" speakers in instructional videos will often reference the literal objects/actions being depicted. We ask: can similar models be trained on more diverse video corpora? And, if so, what types of videos are "grounded" and what types are not? We fit a representative pretraining model to the diverse YouTube8M dataset, and study its success and failure cases. We find that visual-textual grounding is indeed possible across previously unexplored video categories, and that pretraining on a more diverse set results in representations that generalize to both non-instructional and instructional domains.