Spatially-Aware Speaker for Vision-and-Language Navigation Instruction Generation
This work addresses the need for more varied and robust instruction generation in vision-and-language navigation, though it appears incremental as it builds on prior speaker models.
The paper tackles the problem of generating detailed navigational instructions for embodied AI, where existing models produce low-variety instructions and exploit evaluation metrics. The proposed Spatially-Aware Speaker model uses structural and semantic knowledge with adversarial reward learning to outperform existing models on standard metrics.
Embodied AI aims to develop robots that can \textit{understand} and execute human language instructions, as well as communicate in natural languages. On this front, we study the task of generating highly detailed navigational instructions for the embodied robots to follow. Although recent studies have demonstrated significant leaps in the generation of step-by-step instructions from sequences of images, the generated instructions lack variety in terms of their referral to objects and landmarks. Existing speaker models learn strategies to evade the evaluation metrics and obtain higher scores even for low-quality sentences. In this work, we propose SAS (Spatially-Aware Speaker), an instruction generator or \textit{Speaker} model that utilises both structural and semantic knowledge of the environment to produce richer instructions. For training, we employ a reward learning method in an adversarial setting to avoid systematic bias introduced by language evaluation metrics. Empirically, our method outperforms existing instruction generation models, evaluated using standard metrics. Our code is available at \url{https://github.com/gmuraleekrishna/SAS}.