FireNav: Decision-Oriented Reconstruction Evidence Routing for Vision-Language Navigation in Indoor Fire Environments

Hongbo Ding; Xiang Chen; Jiangxin Dong; Jinshan Pan

cs.CVOct 11, 2026

FireNav: Decision-Oriented Reconstruction Evidence Routing for Vision-Language Navigation in Indoor Fire Environments

Authors: Hongbo Ding, Xiang Chen, Jiangxin Dong, Jinshan Pan

Organizations: Nanjing EntroVision Technology Co., Ltd · Nanjing University of Science and Technology

Abstract

In real-world indoor fire environments, visual degradation severely impairs instruction grounding, waypoint prediction, and long-horizon exploration in vision-language navigation (VLN). Although image restoration offers a natural means to mitigate such visual degradation, existing methods primarily optimize reconstruction fidelity, which does not necessarily translate into better navigation decisions and may even corrupt task-critical cues when applied indiscriminately. In this paper, we present FireNav, a decision-oriented reconstruction evidence routing framework that aligns visual restoration with navigation utility in VLN. To advance this task, we first contribute FireVLN-Bench, a paired indoor fire navigation benchmark with clean references preserving task semantics, safety supervision, and matched settings for continuous navigation and robot execution with simulated dynamics. Our FireNav retains native visual tokens and treats latent reconstruction features as auxiliary evidence, where a lightweight router leverages visual features, task context, and prior execution states to determine whether such evidence should be admitted and how strongly it should contribute to the visual tokens. To learn when reconstruction is beneficial, we further introduce decision utility calibration, which compares action prediction losses with and without candidate evidence under the same state, thereby supervising evidence selection according to its contribution to waypoint, heading, and arrival prediction. Experiments demonstrate that FireNav consistently outperforms state-of-the-art baselines in both navigation success and path efficiency across evaluation settings.

Figures & tables

Figure 1: Overview of FireVLN-Bench. The figure shows construction pipeline indoor fire scene dataset statistics and qualitative samples.

Figure 2: Overview of FireNav. A frozen native branch provides visual tokens, while a frozen reconstruction branch exposes mid-level features for fast execution and multiscale features for slow planning. FireNav aligns reconstruction features with native tokens, routes them according to the current decision state, and injects bounded residual evidence while preserving the native visual identity. The slow system uses multiscale evidence for memory-based subgoal planning, whereas the fast system uses mid-level evidence for local waypoint prediction.

Reconstructed RGB Frontends
Method	Venue	Physical Environment				Continuous Environment
Method	Venue	NE $\downarrow$	OS $\uparrow$	SR $\uparrow$	SPL $\uparrow$	NE $\downarrow$	OS $\uparrow$	SR $\uparrow$	SPL $\uparrow$
Native VLN (clean)	-	4.58	68.2	54.1	42.3	3.32	79.7	67.5	56.6
Native VLN (degraded)	-	7.34	43.0	26.8	17.3	5.94	56.3	41.8	29.9
AdaIR ( Cui et al., 2025 )	ICLR’25	7.10	45.1	29.4	19.7	5.82	57.2	42.7	30.5
DCPT ( Hu et al., 2025 )	ICLR’25	6.68	48.8	34.1	25.1	5.33	62.0	47.5	36.6

Table 1: Quantitative comparison on FireVLN-Bench unseen test split. Among methods with degraded observations, best and second-best results are highlighted in bold and underline, respectively.

Figure 3: Qualitative comparison of waypoint predictions and trajectories. Each row shows a model configuration. The first three columns show local waypoint predictions at keyframes, and the last shows trajectories on occupancy maps. Red crosses mark task interruptions ( STOP ).

Variant	Physical Environment			Continuous Environment
Variant	SR $\uparrow$	SPL $\uparrow$	Clean SR $\uparrow$	SR $\uparrow$	SPL $\uparrow$	Clean SR $\uparrow$
Full FireNav	41.6	32.0	52.8	56.1	45.5	67.1
Always-on evidence	36.1	26.0	50.4	50.5	38.6	65.6
w/o decision context	38.9	29.0	53.2	53.7	42.1	67.1
w/o quality statistics	39.8	30.0	53.6	54.4	43.1	66.8
w/o utility calibration	37.8	28.0	52.9	52.2	40.6	66.7

Table 2: Ablation study of FireNav. Clean SR measures retention of the native behavior under clean observations. All variants use the same navigation backbone and evaluation protocol.

Method	Additional Cost		End-to-End Latency		System
Method	Encoder/Restoration (ms) $\downarrow$	Fusion (ms) $\downarrow$	P50 (ms) $\downarrow$	P95 (ms) $\downarrow$	Hz $\uparrow$	Peak GB $\downarrow$
Native VLN	0.0	0.0	280.1	359.2	3.53	18.06
AdaIR	209.9	2.0	491.9	597.4	2.02	20.26
DCPT	96.5	2.0	378.8	464.8	2.62	19.16
MoCE-IR-S	126.1	2.0	408.5	498.1	2.43	19.36
DFPIR	113.9	2.0	396.9	485.0	2.51	19.66

Table 3: Runtime comparison under the same execution setting. P50/P95 denote update latency.

Appendix figures & tables9 assets

Supplementary material from the paper’s appendix.

Appendix

Split	Scenes	Layouts	Routes/Layout/Task	Episodes/Task	Four-Task Episodes
Train	40	160	100	16,000	64,000
Validation Seen	40	160	25	4,000	16,000
Validation Unseen	8	32	25	800	3,200
Test Unseen	12	48	25	1,200	4,800

Appendix

Table 4: Data allocation in FireVLN-Bench. Validation seen uses held-out episodes from the training scenes and therefore does not introduce additional scene identities.

Encoder	Physical Environment			Continuous Environment
Encoder	Always-on	Quality Gate	FireNav	Always-on	Quality Gate	FireNav
AdaIR	23.6	26.0	27.8	35.6	38.0	40.3
DCPT	27.0	28.6	32.5	39.2	41.2	46.0
MoCE-IR-S	26.4	29.4	32.3	39.1	42.1	45.7
DFPIR	25.4	28.6	30.8	37.9	41.2	44.2
R2R-Compos	26.0	29.2	32.0	38.6	41.9	45.5

Appendix

Table 5: Factorial comparison of reconstruction encoders and evidence routing strategies. Values are SPL (%). Each encoder uses the same evidence budget and optimization recipe.

Intervention	Physical Environment			Continuous Environment
Intervention	SR $\uparrow$	SPL $\uparrow$	Clean SR $\uparrow$	SR $\uparrow$	SPL $\uparrow$	Clean SR $\uparrow$
Matched evidence	41.6	32.0	52.8	56.1	45.5	67.1
Exact bypass	26.8	17.3	54.1	41.8	29.9	67.5
Zero spatial features	28.0	18.0	54.0	41.2	29.5	68.1
Shuffled content	29.3	19.0	53.6	42.4	30.4	66.9
Incorrect view content	27.3	17.4	53.4	40.5	29.2	67.3

Appendix

Table 6: Evidence intervention analysis with the same trained FireNav checkpoint. Evidence content or router context is modified without retraining the navigator.

Method	Degraded Observations				Clean Observations
Method	ADE $\downarrow$	FDE $\downarrow$	Heading $\downarrow$	Arrival F1 $\uparrow$	ADE $\downarrow$	Heading $\downarrow$	Arrival F1 $\uparrow$
Native VLN	0.666	1.101	21.0	0.474	0.504	17.0	0.670
MoCE-IR-S	0.595	0.991	18.8	0.546	0.516	17.5	0.657
R2R-Compos	0.615	1.021	19.5	0.523	0.522	17.8	0.644
HOGformer	0.615	1.036	19.8	0.509	0.511	17.3	0.663
BiLaLoRA	0.600	1.001	18.6	0.552	0.514	17.4	0.663

Appendix

Table 7: Action prediction under degraded and clean observations. ADE/FDE are measured in meters, heading error in degrees, and arrival F1 as a fraction.

Method	Image Quality		Action ADE		Cost
Method	PSNR $\uparrow$	SSIM $\uparrow$	Degraded $\downarrow$	Clean $\downarrow$	RGB ms $\downarrow$
Raw degraded	16.8	0.680	0.666	0.504	0.0
AdaIR	15.9	0.648	0.645	0.533	69.9
DCPT	17.8	0.717	0.603	0.518	31.8
MoCE-IR-S	18.2	0.728	0.595	0.516	42.0
DFPIR	17.6	0.711	0.608	0.520	38.1

Appendix

Table 8: Relationship between image reconstruction quality, action prediction, and RGB reconstruction cost.

Method	Physical Environment				Continuous Environment
Method	Collision $\downarrow$	Fall $\downarrow$	Stuck $\downarrow$	False STOP $\downarrow$	Collision $\downarrow$	False STOP $\downarrow$
Native VLN (clean)	10.3	1.9	6.3	3.2	3.5	3.3
Native VLN (degraded)	33.2	3.2	16.2	13.5	19.8	14.5
AdaIR	31.0	3.2	15.7	12.7	18.1	13.2
DCPT	23.6	2.2	13.7	10.0	13.2	10.2
MoCE-IR-S	22.5	2.1	12.8	9.0	11.8	10.1

Appendix

Table 9: Safety and terminal behavior on FireVLN-Bench. All values are percentages of test episodes.

Method / Split	Physical Environment		Continuous Environment
Method / Split	SR $\uparrow$	SPL $\uparrow$	SR $\uparrow$	SPL $\uparrow$
Native VLN / Seen Val.	36.4	26.2	45.8	35.0
Native VLN / Unseen Val.	33.5	22.6	40.0	29.5
Native VLN / Unseen Test	26.8	17.3	41.8	29.9
BiLaLoRA / Seen Val.	44.7	34.0	53.8	43.0
BiLaLoRA / Unseen Val.	40.6	29.6	48.0	37.4

Appendix

Table 10: Generalization across seen validation, unseen validation, and unseen test environments.

Method / Task	Physical Environment		Continuous Environment
Method / Task	SR $\uparrow$	SPL $\uparrow$	SR $\uparrow$	SPL $\uparrow$
Native VLN / Instruction	26.8	17.3	41.8	29.9
Native VLN / Object	20.0	12.0	28.1	18.4
Native VLN / Point	48.3	35.8	63.0	50.7
BiLaLoRA / Instruction	35.6	25.2	48.8	40.1
BiLaLoRA / Object	25.3	16.1	34.3	23.5

Appendix

Table 11: Generalization across instruction, object, and point goal tasks.

Configuration	Cache		Runtime		PE		CE
Configuration	Calls	Hit (%)	P50	P95	SR	SPL	SR	SPL
No history, 3 current views	3	0.0	206.1	256.5	37.9	28.0	52.5	41.5
4 history + 3 current views	3	57.1	242.1	302.2	40.8	30.7	54.8	43.9
10 history + 3 current views	3	76.9	277.9	348.9	41.2	31.5	56.0	45.2
20 history + 3 current views	3	87.0	310.1	389.4	41.6	32.0	56.1	45.5
20 history, front-view evidence only	1	95.2	292.1	369.5	39.2	29.2	54.1	42.8

Appendix

Table 12: Effect of visual history and feature caching.

Vision-Language Navigation

Vision-Language Navigation studies how embodied agents use natural-language instructions and visual observations to plan and execute routes.

See all 310 papers

ReInsVLN: Revisiting Historical Evidence for Training-Free Vision-and-Language Navigation
Oct 19, 2026Siwei Cheng, Hao Wu, Zitong Qi +7Vision-Language Navigation Memory-Augmented VLMs
University of Science and Technology of China · Eastern Institute of Technology, Ningbo · Shanghai Jiao Tong University
FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation
Sep 26, 2026Khang H. Nguyen, Hoang Pham Quang Nguyen, Ha Phuong Nguyen +6Vision-Language Navigation
VinRobotics · University of California, Los Angeles (UCLA) · Center for AI Research, VinUniversity, Vietnam +4
GuidedNav: Shaping Vision-Language Navigation Representations through Pre-action Attention and Spatial Guidance
Oct 9, 2026Jingyu Guo, Jiaxin Huang, Chengxing Lin +6Vision-Language Navigation Visual Navigation
University of Melbourne · AgiBot · Mohamed bin Zayed University of Artificial Intelligence +3

Oct 19, 2026J/K move · Enter open · S save

ReInsVLN: Revisiting Historical Evidence for Training-Free Vision-and-Language Navigation

Siwei Cheng, Hao Wu, Zitong Qi, Xinming Zhang+6

University of Science and Technology of China · Eastern Institute of Technology, Ningbo · Shanghai Jiao Tong University

Vision-Language Navigation Memory-Augmented VLMs VLN Benchmarks

Open paper AskPDF arXiv

Vision-and-language navigation (VLN) relies on retaining and using relevant observations. Yet growing histories introduce redundancy, and retained information may remain underused. We introduce ReInsBench, a diagnostic benchmark that examines history utilization through controlled continuation trials. Its results show that history-review reinstruction can recover navigation failures using the existing task context. Motivated by this finding, we propose ReInsVLN, a training-free framework supporting memory construction and utilization in frozen navigators. Coverage- and Instruction-Aware Compression constructs compact visual context through temporal redundancy reduction, coverage-aware protection, and instruction-guided token selection. Forgetting-Inspired History Reminders guide navigators to revisit retained observations according to a predefined schedule, without additional model calls. On R2R Val-Unseen, ReInsVLN improves success rates by 8.48 and 3.37 percentage points over the native StreamVLN and InternVLA-N1, reaching 64.27% and 64.44%, respectively. Visual compression largely preserves navigation performance while reducing prefill latency and KV cache memory by up to 21.77% and 32.25%, respectively, on matched observation workloads. Real-world trials on humanoid and quadruped robots further support the applicability of ReInsVLN to sequential instruction following and ordinal navigation.

Abstract

Vision-Language Navigation

Explore similar work

ReInsVLN: Revisiting Historical Evidence for Training-Free Vision-and-Language Navigation

FINE: Future-Informed Navigation Encoding for Data-Efficient Vision-Language Navigation

GuidedNav: Shaping Vision-Language Navigation Representations through Pre-action Attention and Spatial Guidance