Multimodal Subtask Graph Generation from Instructional Videos
This addresses the challenge of understanding task structures from noisy web videos, which is incremental as it builds on prior approaches with specific gains.
The paper tackled the problem of modeling causal dependencies between subtasks in instructional videos, presenting MSG2 which generates subtask graphs closer to human annotations and improves next subtask prediction accuracy by 85% and 30% on ProceL and CrossTask datasets.
Real-world tasks consist of multiple inter-dependent subtasks (e.g., a dirty pan needs to be washed before it can be used for cooking). In this work, we aim to model the causal dependencies between such subtasks from instructional videos describing the task. This is a challenging problem since complete information about the world is often inaccessible from videos, which demands robust learning mechanisms to understand the causal structure of events. We present Multimodal Subtask Graph Generation (MSG2), an approach that constructs a Subtask Graph defining the dependency between a task's subtasks relevant to a task from noisy web videos. Graphs generated by our multimodal approach are closer to human-annotated graphs compared to prior approaches. MSG2 further performs the downstream task of next subtask prediction 85% and 30% more accurately than recent video transformer models in the ProceL and CrossTask datasets, respectively.