Abstract
Adapting vision-language navigation (VLN) policies to new environments is expensive because every additional route and instruction requires an embodied demonstration. Yet standard observation-to-action training uses only a small fraction of the information already contained in each trajectory. In particular, future observations reveal the instruction-relevant landmarks that the agent will encounter, including what they look like and how they are arranged in 3D. We introduce FINE, a Future-Informed Navigation Encoding framework that extracts this latent supervision from existing demonstrations. FINE equips a VLN backbone with two complementary auxiliary representations. First, explicit landmark tokens follow the ordered landmarks specified by the instruction and are trained to predict future landmark regions in both semantic 2D patch-feature space and viewpoint-dependent 3D geometric feature space. Second, an implicit future token learns to distinguish the landmark state that is actually reached from plausible same-scene counterfactual futures generated by a video world model. On R2R-CE and RxR-CE val-unseen, FINE improves InternVLA-N1 by 2.6 and 4.5 success-rate points, respectively, at full training data. More importantly, as demonstrations become limited, the benefit grows: at a 70% demonstration budget, FINE improves success rate by 6.8 points, recovering roughly one-third of the performance lost by reducing the training demonstrations. Project page is available at https://finevln.github.io/.
Keywords
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Nguyen, K. H., Nguyen, H. P. Q., Nguyen, H. P., Binh, K. D., Nguyen, X. H., Ngo, V., Minh, D. H. N., Nguyen, H., & Le, A. T. (2026). FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation. https://omanscience.com/en/articles/fine-future-informed-navigation-encoding-for-data-efficient-vision-language-navigation
MLA 9
Nguyen, Khang H., et al. "FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation." https://omanscience.com/en/articles/fine-future-informed-navigation-encoding-for-data-efficient-vision-language-navigation.
Chicago (author–date)
Nguyen, Khang H., Hoang Pham Quang Nguyen, Ha Phuong Nguyen, Khanh Dinh Binh, Xuan Ha Nguyen, Vien Ngo, Duy Ho Nguyen Minh, Huan Nguyen, and An T. Le. 2026. "FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation." https://omanscience.com/en/articles/fine-future-informed-navigation-encoding-for-data-efficient-vision-language-navigation.
Harvard
Nguyen, K. H., Nguyen, H. P. Q., Nguyen, H. P., Binh, K. D., Nguyen, X. H., Ngo, V., Minh, D. H. N., Nguyen, H. and Le, A. T. (2026) 'FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation', Available at: https://omanscience.com/en/articles/fine-future-informed-navigation-encoding-for-data-efficient-vision-language-navigation.
Vancouver
Nguyen KH, Nguyen HPQ, Nguyen HP, Binh KD, Nguyen XH, Ngo V, et al. FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation. https://omanscience.com/en/articles/fine-future-informed-navigation-encoding-for-data-efficient-vision-language-navigation
IEEE
K. H. Nguyen, H. P. Q. Nguyen, H. P. Nguyen, K. D. Binh, X. H. Nguyen, V. Ngo, D. H. N. Minh, H. Nguyen, and A. T. Le, "FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation," https://omanscience.com/en/articles/fine-future-informed-navigation-encoding-for-data-efficient-vision-language-navigation.