AI Paper Atlas

Research map / Benchmarks / R2R val unseen

R2R val unseen leaderboard

67 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on R2R val unseen. Best success rate: 84.4 (SRDF round 3, ICLR 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

R2R val unseen · success rate

60 papers, each with its best reported number for its own method; higher is better.

#success rate ↑MethodPaperTopic
184.4SRDF round 3
DUET
Bootstrapping Language-Guided Navigation Learning with Self-Refining Data Flywheel
ICLR 2025 · Table: Navigator and instructi
Robot action and navigation
281DUET*+ScaleVLN (EnvEdit, CLIP ViT-H/14)
DUET
Scaling Data Generation in Vision-and-Language Navigation
ICCV 2023 · Table 9
Robot action and navigation
380.7MC (ε=0.5)Just Ask: An Interactive Learning Framework for Vision and Language Navigation
AAAI 2020 · Table 1
Human-AI and goal-conditioned
478.88MAGIC-LMAGIC: Meta-Ability Guided Interactive Chain-of-Distillation for Effective-and-Efficient Vision-and-Language Navigation
TPAMI 2026 · Table 1
Knowledge distillation
578Ours
DUET
Uncertainty-Aware Gaussian Map for Vision-Language Navigation
ICLR 2026 · Table 2
Robot action and navigation
677R3 (ours)Run, Ruminate, and Regulate: A Dual-process Thinking System for Vision-and-Language Navigation
AAAI 2026 · Table 1
Multimodal LLMs
7773D Gaussian Map (Ours)3D Gaussian Map with Open-Set Semantic Grouping for Vision-Language Navigation
ICCV 2025 · Table 2
Gaussian splatting
876+IDEA
DUET
Turning Adaptation into Assets: Cross-Domain Bridging for Online Vision-Language Navigation
ICML 2026 · Table 2
Robot action and navigation
975.8VER neighborhood range 9Volumetric Environment Representation for Vision-Language Navigation
CVPR 2024 · Table 5
3D generation
1075.01Airbert + Speaker + Follower ensembleAirbert: In-Domain Pretraining for Vision-and-Language Navigation
ICCV 2021 · Table 5
Robot action and navigation
50 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

R2R val unseen · SPL

41 papers, each with its best reported number for its own method; higher is better.

#SPL ↑MethodPaperTopic
178SALIPlanning from Imagination: Episodic Simulation and Episodic Memory for Vision-and-Language Navigation
AAAI 2025 · Table 1
Human-AI and goal-conditioned
269.3GOAT + ATENA
GOAT
Active Test-time Vision-Language Navigation
NeurIPS 2025 · Table: Experimental results on
Human-AI and goal-conditioned
366Ours
DUET
Uncertainty-Aware Gaussian Map for Vision-Language Navigation
ICLR 2026 · Table 2
Robot action and navigation
466R3 (ours)Run, Ruminate, and Regulate: A Dual-process Thinking System for Vision-and-Language Navigation
AAAI 2026 · Table 1
Multimodal LLMs
5663D Gaussian Map (Ours)3D Gaussian Map with Open-Set Semantic Grouping for Vision-Language Navigation
ICCV 2025 · Table 2
Gaussian splatting
665.16directed path representation (full)
DINOv2
PRET: Planning with Directed Fidelity Trajectory for Vision and Language Navigation
ECCV 2024 · Table 4
Robot action and navigation
765DUET w/ FEEDTTA
DUET
Test-Time Adaptation for Online Vision-Language Navigation with Feedback-based Reinforcement Learning
ICML 2025 · Table 2
Robot action and navigation
865GELA
HAMT
Grounded Entity-Landmark Adaptive Pre-Training for Vision-and-Language Navigation
ICCV 2023 · Table 3
Human and embodied modeling
964.8SUSAAgent Journey Beyond RGB: Hierarchical Semantic-Spatial Representation Enrichment for Vision-and-Language Navigation
AAAI 2026 · Table 1
Robot action and navigation
1064.3PanoGen
DUET-CLIP
PanoGen: Text-Conditioned Panoramic Environment Generation for Vision-and-Language Navigation
NeurIPS 2023 · Table 1
Robot action and navigation
31 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

R2R val unseen · NE

12 papers, each with its best reported number for its own method; lower is better.

2 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.