Correctable Landmark Discovery via Large Models for Vision-Language Navigation
IEEE Transactions on Pattern Analysis and Machine Intelligence · 2024-12 · 46 (12) · 8534-8548
DOI: 10.1109/tpami.2024.3407759 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024-12. Verfasst von Bingqian Lin und 7 weiteren Autor:innen. Die Arbeit wurde bisher 31-mal zitiert (Crossref) und nennt 58 Literaturangaben. Gefördert von: National Science and Technology Major Project, Guangdong Outstanding Youth Fund, Shenzhen Science and Technology Program.
Abstract (Originalsprache)
Vision-Language Navigation (VLN) requires the agent to follow language instructions to reach a target position. A key factor for successful navigation is to align the landmarks implied in the instruction with diverse visual observations. However, previous VLN agents fail to perform accurate modality alignment especially in unexplored scenes, since they learn from limited navigation data and lack sufficient open-world alignment knowledge. In this work, we propose a new VLN paradigm, called COrrectable LaNdmark DiScOvery via Large ModEls (CONSOLE). In CONSOLE, we cast VLN as an open-world sequential landmark discovery problem, by introducing a novel correctable landmark discovery scheme based on two large models ChatGPT and CLIP. Specifically, we use ChatGPT to provide rich open-world landmark cooccurrence commonsense, and conduct CLIP-driven landmark discovery based on these commonsense priors. To mitigate the noise in the priors due to the lack of visual constraints, we introduce a learnable cooccurrence scoring module, which corrects the importance of each cooccurrence according to actual observations for accurate landmark discovery. We further design an observation enhancement strategy for an elegant combination of our framework with different VLN agents, where we utilize the corrected landmark features to obtain enhanced observation features for action decision. Extensive experimental results on multiple popular VLN benchmarks (R2R, REVERIE, R4R, RxR) show the significant superiority of CONSOLE over strong baselines. Especially, our CONSOLE establishes the new state-of-the-art results on R2R and R4R in unseen scenarios.
Förderung
- National Science and Technology Major Project
- Guangdong Outstanding Youth Fund
- Shenzhen Science and Technology Program
- Nansha Key R&D Program
- CAAI-Huawei MindSpore Open Fund
Zitieren
APA
Lin, B., Nie, Y., Wei, Z., Zhu, Y., Xu, H., Ma, S., Liu, J., & Liang, X. (2024). Correctable Landmark Discovery via Large Models for Vision-Language Navigation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(12), 8534-8548. https://doi.org/10.1109/tpami.2024.3407759
Vancouver
Lin B, Nie Y, Wei Z, Zhu Y, Xu H, Ma S, et al. Correctable Landmark Discovery via Large Models for Vision-Language Navigation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2024;46(12):8534-8548. doi:10.1109/tpami.2024.3407759
BibTeX
@article{Lin2024,
author = {Lin, Bingqian and Nie, Yunshuang and Wei, Ziming and Zhu, Yi and Xu, Hang and Ma, Shikui and Liu, Jianzhuang and Liang, Xiaodan},
title = {Correctable Landmark Discovery via Large Models for Vision-Language Navigation},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
year = {2024},
volume = {46},
number = {12},
pages = {8534--8548},
doi = {10.1109/tpami.2024.3407759}
}
EUVIMED