A Dataset for Interactive Vision-Language Navigation with Unknown Command Feasibility

원문 초록

Vision-language navigation (VLN), in which an agent follows language instruction in a visual environment, has been studied under the premise that the input command is fully feasible in the environment. Yet in practice, a request may not be possible due to language ambiguity or environment changes. To study VLN with unknown command feasibility, we introduce a new dataset Mobile app Tasks with Iterative Feedback (MoTIF), where the goal is to complete a natural language command in a mobile app. Mobile apps provide a scalable domain to study real downstream uses of VLN methods. Moreover, mobile app commands provide instruction for interactive navigation, as they result in action sequences with state changes via clicking, typing, or swiping. MoTIF is the first to include feasibility annotations, containing both binary feasibility labels and fine-grained labels for why tasks are unsatisfiable. We further collect follow-up questions for ambiguous queries to enable research on task uncertainty resolution. Equipped with our dataset, we propose the new problem of feasibility prediction, in which a natural language instruction and multimodal app environment are used to predict command feasibility. MoTIF provides a more realistic app dataset as it contains many diverse environments, high-level goals, and longer action sequences than prior work. We evaluate interactive VLN methods using MoTIF, quantify the generalization ability of current approaches to new app environments, and measure the effect of task feasibility on navigation performance.