LOC-ZSON: Language-driven Object-Centric Zero-Shot Object Retrieval and Navigation
This addresses the challenge of enabling robots to navigate to objects based on language queries in complex environments, representing an incremental advancement in visual-language models for robotics.
The paper tackles the problem of zero-shot object retrieval and navigation in complex scenes by proposing LOC-ZSON, a language-driven object-centric image representation, which improves text-to-image recall by 1.38-13.38% and navigation success rates by 5% in simulation and 16.67% in real-world environments.
In this paper, we present LOC-ZSON, a novel Language-driven Object-Centric image representation for object navigation task within complex scenes. We propose an object-centric image representation and corresponding losses for visual-language model (VLM) fine-tuning, which can handle complex object-level queries. In addition, we design a novel LLM-based augmentation and prompt templates for stability during training and zero-shot inference. We implement our method on Astro robot and deploy it in both simulated and real-world environments for zero-shot object navigation. We show that our proposed method can achieve an improvement of 1.38 - 13.38% in terms of text-to-image recall on different benchmark settings for the retrieval task. For object navigation, we show the benefit of our approach in simulation and real world, showing 5% and 16.67% improvement in terms of navigation success rate, respectively.