Abstract
Embodied agents are expected to perform object navigation in dynamic, open-world environments. However, existing approaches typically rely on static trajectories and a fixed set of object categories during training, overlooking the real-world requirement for continual adaptation to evolving scenarios. To facilitate related studies, we introduce the continual object navigation benchmark, which requires agents to acquire navigation skills for new object categories while avoiding catastrophic forgetting of previously learned knowledge. To tackle this challenge, we propose C-Nav, a continual visual navigation framework that integrates two key innovations: (1) A dual-path anti-forgetting mechanism, which comprises feature distillation that aligns multi-modal inputs into a consistent representation space to ensure representation consistency, and feature replay that retains temporal features within the action decoder to ensure policy consistency. (2) An adaptive sampling strategy that selects diverse and informative experiences, thereby reducing redundancy and minimizing memory overhead. Extensive experiments across multiple model architectures demonstrate that C-Nav consistently outperforms existing approaches, achieving superior performance even compared to baselines with full trajectory retention, while significantly lowering memory requirements.
Continual Object Navigation: The robot must continually learn from new data while retaining its ability to navigate to previously seen object goals.
Different navigation model architectures. According to the action decoder, they can be categorized into a) RNN-Based, b) Bev-Based, c) Transformer-Based, and d) LLM-Based. After training across multiple phases, the baseline model suffers from representation drift and policy degradation, leading to a loss of navigation capability on previously learned tasks
C-Nav Framework: Overview of the proposed framework with adaptive experience selection and dual-path anti-forgetting mechanism.
Case Study
Side-by-side comparison of Baseline model and C-Nav model navigation performance
Baseline Model
Baseline model showing catastrophic forgetting and navigation failures
C-Nav Model
C-Nav model with continual learning and successful navigation
Additional Case Study
BibTeX
@article{CNav2025,
title={C-NAV: Towards Self-Evolving Continual Object Navigation in Open World},
author={Ming-Ming Yu and Fei Zhu and Wenzhuo Liu and Yirong Yang and Qunbo Wang and Wenjun Wu and Jing Liu},
journal={Advances in Neural Information Processing Systems (NeurIPS)},
year={2025},
}
}