From Instruction Following to Cognitive Navigation: A Survey on the Evolution of Vision-and-Language Navigation
Escrito por Kailin Lyu, Kangyi Wu, Pengna Li, Wenxuan Song, Di Wu, Jianwei He, Junting Chen, Ning Yang, Zebin Han, Kaiwen Luo, Liang Lin, Long Xiao, Xi Lin, Weigang Xue, Yongen Zhao, Kaiwen Xue, Zhiqiang Yuan, Yang Liu, WeiZhong Huang, Liwei Yang, Jinwei He, Nanxing Hu, Jinjun Wang, Ye Deng, Shaoqing Xu, Lin Zhao, Shengqian Qin, Fan Xu, Qingyi Si, Keji He, Jiaqi Peng, Hao Wu, Hao Chen, Xiaoyu Ma, Zhenghong Zhou, Zeqin Liao, Qiankun Li, Kun Wang, Ce Hao, Lianyu Hu, Dongrui Liu, Chuang Zhu, Yonggang Qi, Xingjun Ma, Hao Chen, Qing Guo, Weinan Zhang, Qi Li, Shanghang Zhang, Zhigang Zeng, Chunhua Shen, Yu-Gang Jiang, Tieniu Tan y Yang Liu
- Publicado
- Servidor
- Preprints.org
Vision-and-Language Navigation (VLN) requires embodied agents to ground natural language instructions in visual perception and make navigation decisions in complex 3D environments, making it a central problem in embodied artificial intelligence. Since the introduction of the Room-to-Room (R2R) benchmark, VLN has made substantial progress. In recent years, as research settings have gradually expanded from closed and single indoor benchmark scenarios to open-world environments, the field has undergone a profound paradigm shift from passive instruction following on fixed benchmarks to autonomous cognitive navigation in open-world settings. However, existing surveys mainly organize prior work according to technical taxonomies, lacking a systematic characterization of this paradigm evolution. To address this gap, this survey proposes an evolution-centered unified analytical framework that reviews contemporary VLN research across four progressive layers: perception, cognition, learning, and generalization. It reveals the intrinsic connections and evolutionary logic among different technical lines, identifies key open challenges at each dimension, and outlines future research directions. This survey aims to provide VLN researchers with a clear panoramic view of capability evolution, while offering the broader embodied intelligence community a systematic roadmap from closed-benchmark evaluation toward trustworthy open-world deployment.