Learning to Segment Actions from Observation and Narration
This addresses the problem of segmenting actions in videos without full supervision, which is incremental as it builds on existing methods by incorporating narrative language.
The paper tackles unsupervised and weakly-supervised action segmentation in video by using a generative segmental model guided by narration, achieving competitive performance on a dataset of instructional videos.
We apply a generative segmental model of task structure, guided by narration, to action segmentation in video. We focus on unsupervised and weakly-supervised settings where no action labels are known during training. Despite its simplicity, our model performs competitively with previous work on a dataset of naturalistic instructional videos. Our model allows us to vary the sources of supervision used in training, and we find that both task structure and narrative language provide large benefits in segmentation quality.