Towards Learning a Generalist Model for Embodied Navigation

4 Dec 2023  ยท  Duo Zheng, Shijia Huang, Lin Zhao, Yiwu Zhong, LiWei Wang ยท

Building a generalist agent that can interact with the world is the intriguing target of AI systems, thus spurring the research for embodied navigation, where an agent is required to navigate according to instructions or respond to queries. Despite the major progress attained, previous works primarily focus on task-specific agents and lack generalizability to unseen scenarios. Recently, LLMs have presented remarkable capabilities across various fields, and provided a promising opportunity for embodied navigation. Drawing on this, we propose the first generalist model for embodied navigation, NaviLLM. It adapts LLMs to embodied navigation by introducing schema-based instruction. The schema-based instruction flexibly casts various tasks into generation problems, thereby unifying a wide range of tasks. This approach allows us to integrate diverse data sources from various datasets into the training, equipping NaviLLM with a wide range of capabilities required by embodied navigation. We conduct extensive experiments to evaluate the performance and generalizability of our model. The experimental results demonstrate that our unified model achieves state-of-the-art performance on CVDN, SOON, and ScanQA. Specifically, it surpasses the previous stats-of-the-art method by a significant margin of 29% in goal progress on CVDN. Moreover, our model also demonstrates strong generalizability and presents impressive results on unseen tasks, e.g., embodied question answering and 3D captioning.

PDF Abstract

Datasets


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Visual Navigation Cooperative Vision-and-Dialogue Navigation NaviLLM dist_to_end_reduction 7.90 # 1
spl 0.09 # 14
Visual Navigation R2R NaviLLM spl 0.60 # 2
3D Question Answering (3D-QA) ScanQA Test w/ objects NaviLLM Exact Match 26.27 # 2
BLEU-1 39.73 # 1
BLEU-4 13.90 # 2
ROUGE 40.23 # 2
METEOR 16.56 # 1
CIDEr 80.77 # 2
Visual Navigation SOON Test NaviLLM Nav-SPL 26.26 # 2
SR 35.04 # 3

Methods