Visual Goal-Step Inference using wikiHow
This work addresses the challenge of multimodal reasoning about procedural events for AI systems, though it is incremental as it adapts a text-based task to the visual domain.
The paper tackles the problem of visual goal-step inference by introducing a new task where a model must select an image representing a plausible step towards a textual goal, using a dataset of 772,277 images from wikiHow, and shows that learned representations transfer to other datasets, increasing accuracy by 15-20%.
Understanding what sequence of steps are needed to complete a goal can help artificial intelligence systems reason about human activities. Past work in NLP has examined the task of goal-step inference for text. We introduce the visual analogue. We propose the Visual Goal-Step Inference (VGSI) task, where a model is given a textual goal and must choose which of four images represents a plausible step towards that goal. With a new dataset harvested from wikiHow consisting of 772,277 images representing human actions, we show that our task is challenging for state-of-the-art multimodal models. Moreover, the multimodal representation learned from our data can be effectively transferred to other datasets like HowTo100m, increasing the VGSI accuracy by 15 - 20%. Our task will facilitate multimodal reasoning about procedural events.