VideoBERT: A Joint Model for Video and Language Representation Learning

Self-supervised learning has become increasingly important to leverage the abundance of unlabeled data available on platforms like YouTube. Whereas most existing approaches learn low-level representations, we propose a joint visual-linguistic model to learn high-level features without any explicit supervision. In particular, inspired by its recent success in language modeling, we build upon the BERT model to learn bidirectional joint distributions over sequences of visual and linguistic tokens, derived from vector quantization of video data and off-the-shelf speech recognition outputs, respectively. We use VideoBERT in numerous tasks, including action classification and video captioning. We show that it can be applied directly to open-vocabulary classification, and confirm that large amounts of training data and cross-modal information are critical to performance. Furthermore, we outperform the state-of-the-art on video captioning, and quantitative results verify that the model learns high-level semantic features.

PDF Abstract ICCV 2019 PDF ICCV 2019 Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Video Captioning YouCook2 VideoBERT + S3D BLEU-3 7.59 # 6
BLEU-4 4.33 # 11
METEOR 11.94 # 8
ROUGE-L 28.80 # 9
CIDEr 0.55 # 10
Action Classification YouCook2 VideoBERT (cross modal) Object Top 5 Accuracy 33.7 # 1
Object Top-1 Accuracy 13.1 # 1
Verb Top-1 Accuracy 3.2 # 1
Verb Top-5 Accuracy 43.3 # 1

Methods