Vision-Language Navigation on R2R Unseen (test)

86SRHuman

Evaluation Results

Method
Human2022.04	86	—	1.61	0.76	—	11.85
Human2023.03	86	—	1.61	0.76	—	11.85
SRDF2025.08	84	—	1.88	0.77	88	—
SkillNav (SRDF-Aug)2025.08	84	—	1.76	0.77	87	—
SkillNav (ScaleVLN-Aug)2025.08	78	—	2.53	0.7	83	—
Airbert2021.08	77	687	2.69	0.01	99	—
ScaleVLN2025.08	77	—	2.73	0.68	84	—
GOAT2024.04	74.57	—	3.04	0.6494	80.35	—
BTK2026.03	74	—	3.07	0.63	81	—
SAME2025.08	74	—	3.03	0.64	—	—
VLN-BERT2020.04	73	687	3.09	0.01	99	—
GridMM2024.04	73	—	3.35	0.62	—	—
BEVBert2024.04	73	—	3.13	0.62	81	—
GridMM2024.12	73	—	—	0.62	—	—
VLN-BERT2021.08	73	687	3.09	0.01	99	—
GridMM2026.03	73	—	3.35	0.62	—	—
BEVBert2026.03	73	—	3.13	0.62	81	—
BSG2026.03	73	—	3.19	0.62	—	—
BEVBERT2025.08	73	—	3.13	0.62	81	—
SUSA2024.12	72.5	—	3.2	0.638	—	—
DSRG2024.04	72	—	3.33	0.61	78	—
CONSOLE2024.12	72	—	3.3	0.61	—	—
NavGPT22026.03	72	—	3.33	0.6	80	—
KESU2026.03	72	—	3.31	0.61	—	—
CONSOLE2026.03	72	—	3.3	0.61	—	—
TSOR2026.03	72	—	3.27	0.62	—	—
NavGPT-2 (FlanT5-5B)2025.08	72	—	3.33	0.6	80	—
Auxiliary Tasks2020.04	71	41	3.24	0.21	81	—
Meta-Explore2023.03	71	—	3.57	0.61	—	14.25
AuxRN2021.08	71	41	3.24	0.21	81	—
ACME2024.12	70.4	—	3.68	0.612	—	—
KERM2026.03	70	—	3.61	0.59	—	—
ViTeC2026.03	70	—	3.31	0.6	—	—
KERM2023.03	69.73	—	3.61	0.5925	—	14.6
KERM2024.12	69.7	—	3.6	0.592	—	—
DUET2023.03	69.25	—	3.65	0.5868	—	14.74
Environmental Dropout2020.04	69	687	3.26	0.01	99	—
DUET2023.03	69	—	3.65	0.59	—	14.73
DUET2023.03	69	—	3.65	0.59	—	14.73
DUET2024.04	69	—	3.65	0.59	76	—
AZHP2024.12	69	—	3.4	0.59	—	—
FDA2024.12	69	—	3.41	0.62	—	—
DUET2024.12	69	—	3.65	0.59	—	—
EnvDrop2021.08	69	687	3.26	0.01	99	—
DUET2026.03	69	—	3.65	0.59	76	—
DUET2025.08	69	—	3.65	0.59	76	—
EnvEdit2024.04	68	—	3.59	0.64	—	—
NaviLLM (Vicuna-7B)2025.08	68	—	3.71	0.6	—	—
TD-STP2024.04	67	—	3.73	0.61	72	—
GELA2024.04	67	—	3.59	0.62	—	—
TD-STP2026.03	67	—	3.73	0.61	72	—
HOP+2026.03	66	—	3.71	0.6	—	—
Esceme2026.03	66	—	3.77	0.63	—	—
VLNBERT+REM2022.04	65	—	3.87	0.59	—	13.11
HAMT-e2e2023.03	65	—	3.93	0.6	—	12.27
HAMT2023.03	65	—	3.93	0.6	—	12.27
HAMT2024.04	65	—	3.93	0.6	72	—
GeoVLN2024.04	65	—	3.95	0.61	—	—
HAMT2026.03	65	—	3.93	0.6	72	—
HAMT2025.08	65	—	3.93	0.6	72	—
HOP (1)2022.03	64	—	3.87	0.58	—	13.29
HOP (2)2022.03	64	—	3.83	0.59	—	12.68
HOP2023.03	64	—	3.83	0.59	—	12.68
VLN-PETL2023.08	63.22	—	4.1	0.5825	—	12.3
Fine-Tuning2023.08	63.2	—	4.09	0.5855	—	12.2
Reinforced Cross-Modal Matching2020.04	63	358	4.03	0.02	96	—
RecBERT2022.03	63	—	4.09	0.57	—	12.35
VLNBERT2022.04	63	—	4.09	0.57	—	12.35
VLNBERT2023.03	63	—	4.09	0.57	—	12.35
VLNBERT2023.03	63	—	4.09	0.57	—	12.35
Reinforced CM2021.08	63	358	4.03	0.02	96	—
LSAL2026.03	63	—	4	0.58	—	—
AirBERT2022.03	62	—	4.13	0.57	—	12.41
NvEM+SEvol2022.04	62	—	4.13	0.57	—	13.4
AirBERT2023.03	62	—	4.13	0.67	—	12.41
LORA2023.08	61.44	—	4.15	0.5596	—	12.99
Tactical Rewind2020.04	61	197	4.29	0.03	90	—
Self-Monitoring2020.04	61	373	4.48	0.02	97	—
SSM2022.04	61	—	4.57	0.46	—	20.4
SMNA2023.03	61	—	4.48	0.56	—	—
FAST2023.03	61	—	4.29	0.63	—	196.53
SSM2023.03	61	—	4.57	0.46	—	20.4
Self-Monitoring2021.08	61	373	4.48	0.02	97	—
Adapter2023.08	60.69	—	4.27	0.5588	—	13.19
ActiveVLN2022.03	60	—	4.33	0.41	—	21.6
NvEM+SEvol2022.04	60	—	4.25	0.55	—	13.1
EnvDrop+SEvol2022.04	59	—	3.7	0.55	—	14.3
PREVALENT2021.08	59	10.21	4.52	0.56	64	—
NvEM2022.03	58	—	4.37	0.54	—	12.98
NvEM2022.04	58	—	4.37	0.54	—	12.98
BitFit2023.08	57.15	—	4.63	0.5303	—	12.96
VLNBERT2022.04	57	—	4.59	0.53	—	12.34
ORIST2022.04	57	—	5.1	0.52	—	11.31
SSM2022.04	57	—	4.66	0.44	—	18.5
EnvDrop+SEvol2022.04	57	—	4.52	0.52	—	15.48
Active Perception2022.04	56	—	4.77	0.37	—	21
EntityGraph2022.03	55	—	4.75	0.52	—	10.29
HOP (0)2022.03	55	—	4.93	0.5	—	12.53
AuxRN2022.04	55	—	5.15	0.51	—	—
RelGraph2022.04	55	—	4.75	0.52	—	10.29