Robustness of Utilizing Feedback in Embodied Visual Navigation
This addresses robustness in embodied AI navigation for scenarios with intermittent human assistance, but it is incremental as it builds on existing feedback-based methods.
The paper tackles the problem of training an agent to request help for object-goal navigation, using a curriculum with mixed feedback episodes to enhance robustness when a teacher is unavailable, resulting in improved performance without feedback.
This paper presents a framework for training an agent to actively request help in object-goal navigation tasks, with feedback indicating the location of the target object in its field of view. To make the agent more robust in scenarios where a teacher may not always be available, the proposed training curriculum includes a mix of episodes with and without feedback. The results show that this approach improves the agent's performance, even in the absence of feedback.