Pith. sign in

REVIEW 4 major objections 2 minor 11 cited by

A navigation model trained on its own error trajectories sets new records on two vision-and-language benchmarks.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

By iteratively retraining on automatically generated corrective examples derived from its own wrong paths, CorrectNav reports new state-of-the-art success rates of 65.1% (R2R-CE) and 69.3% (RxR-CE).

T0 review reviewed 2026-08-05 challenge →

load-bearing objection Flywheel idea is worth a serious referee, but the supplied full text is a different paper—the SOTA claims are unverifiable from the abstract alone. the 4 major comments →

arxiv 2508.10416 v1 pith:JAKPY3YF submitted 2025-08-14 cs.RO cs.AIcs.CLcs.CV

CorrectNav: Self-Correction Flywheel Empowers Vision-Language-Action Navigation Model

Zhuoyuan Yu , Yuxing Long , Zihan Yang , Chengyan Zeng , Hongwei Fan , Jiyao Zhang , Hao Dong This is my paper
classification cs.RO cs.AIcs.CLcs.CV
keywords vision-and-language navigationself-correctionerror trajectoriespost-trainingVLA modelR2R-CERxR-CErobotic navigation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that a vision-language-action (VLA) navigation model can improve itself by treating its own wrong turns on the training set as a valuable resource instead of discarding them. It introduces the Self-correction Flywheel, a post-training loop that detects where the model deviates, automatically generates corrective perception and action labels for those deviation points, and retrains on them. Re-running the loop repeatedly exposes new errors, each round creating fresh training data. The resulting monocular RGB-based model, CorrectNav, is reported to reach 65.1% success on R2R-CE and 69.3% on RxR-CE, surpassing prior VLA navigation models by 8.2% and 16.4%, and to recover from its own errors on real robots. If true, this would show that autonomous self-correction data from a model's own failures can substitute for expensive human corrections in instruction-following navigation.

Core claim

The central claim is that a VLA navigation model can bootstrap its own error-correction ability through a closed loop: run the model on the training set, detect points where it deviates from the ground-truth path, auto-generate two kinds of corrective data—perception corrections for what the model should have attended to or recognized, and action corrections for what it should have done—then fine-tune on those samples; after retraining, evaluate again on the training set to find newly exposed errors, and repeat. Applying this flywheel to a monocular RGB VLA model yields, the authors report, state-of-the-art success rates of 65.1% on R2R-CE and 69.3% on RxR-CE, improvements of 8.2% and 16.4%

What carries the argument

The Self-correction Flywheel is the central mechanism: an iterative post-training loop that (1) runs the current model on the training split, (2) uses a deviation detector to locate segments where the agent leaves the correct path, (3) synthesizes self-correction data for both perception and action at those deviation points, and (4) fine-tunes the model on that data; the loop repeats, with each pass's errors becoming the next pass's training fuel. The design insight is that error trajectories are not waste but a renewable data source that improves the model's ability to recognize and recover from its own mistakes.

Load-bearing premise

The load-bearing premise is that the automatic deviation detector and the auto-generated self-correction labels, for perception and action, are accurate enough to serve as a reliable training signal, and that repeated re-evaluation on the training set (the flywheel's fuel) never contaminates the benchmark splits behind the reported numbers.

What would settle it

A direct check is to rerun the flywheel with the deviation detector and label-generation modules disabled—training only on the same original data without self-generated corrections—and verify that the reported success-rate gains on R2R-CE and RxR-CE disappear; if the gains persist, the flywheel's specific mechanism is not the cause. Alternatively, one can inspect the generated self-correction labels for hallucinated actions or perceptions on a random sample of error trajectories and compute agreement with human judgments; low agreement would indicate the training signal is noisy.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • If the flywheel works as reported, other instruction-following agents—grounded language understanding, manipulation, or driving—could adopt the same self-correction loop to improve without new human annotations.
  • The method implies that error detection itself can be learned automatically from the model's own deviations, reducing the need for external supervision for correction.
  • The reported gains on R2R-CE and RxR-CE suggest that self-generated corrections transfer to unseen environments from the same benchmarks, and the real-robot results suggest transfer to physical platforms.
  • Retraining on the training set's error trajectories is a form of hard-example mining; the flywheel's iteration count becomes a new hyperparameter that trades compute against performance.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • A testable extension is to apply the same flywheel on unlabeled or weakly labeled trajectories, since the current loop's deviation detector relies on ground-truth path comparisons that may not exist in novel environments.
  • If the flywheel's gains come mostly from perception corrections, ablating the action-correction stream should drastically reduce the improvement; this isolates whether self-correction is primarily perceptual or behavioral.
  • The repeated re-evaluation on the training set raises the risk of overfitting to training-set pathologies; reporting per-iteration performance on a held-out validation split would show whether gains are monotonic or saturate.
  • Combining the generated corrective trajectories with data augmentation on instructions or viewpoints could produce more robust self-correction signals, though the paper does not test this.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 2 minor

Summary. The submission, as supplied, consists of an abstract for "CorrectNav: Self-Correction Flywheel Empowers Vision-Language-Action Navigation Model" followed by the full text of an unrelated COLM 2025 paper, "Evaluating LLMs on Chinese Idiom Translation" by Yang et al. The abstract proposes a Self-correction Flywheel post-training paradigm: the model's own error trajectories on the training set are used to automatically generate self-correction data for perception and action, the model is retrained on this data, and the process is repeated over multiple flywheel iterations. The abstract reports new state-of-the-art success rates of 65.1% on R2R-CE and 69.3% on RxR-CE, surpassing prior VLA navigation models by 8.2% and 16.4%, plus qualitative real-robot tests. No methods, experiments, baselines, tables, or implementation details for CorrectNav appear anywhere in the manuscript.

Significance. If substantiated, the iterative self-correction loop from automatically generated labels would be a meaningful post-training paradigm for embodied navigation, potentially reducing the need for human annotation and improving error recovery. The reported gains over prior VLA models are large. However, because the submitted full text is not the CorrectNav paper, none of the supporting evidence can be checked. As submitted, the contribution is an abstract-level claim with no verifiable scientific content.

major comments (4)
  1. [Full text (pp. 1-24)] The body of the submission is arXiv:2508.10421, "Evaluating LLMs on Chinese Idiom Translation," which has no overlap in title, authors, or topic with CorrectNav. There is no description of the deviation detector, the automatic self-correction data generation for perception and action, the model architecture, the flywheel iteration procedure, the benchmark evaluation protocol, or the real-robot setup. The central state-of-the-art claim in the abstract therefore has no supporting evidence in the manuscript. This is a load-bearing failure that cannot be remedied by local revision; the correct manuscript must be supplied and then reviewed in full.
  2. [Abstract] The abstract states that the flywheel "re-evaluates the model on the training set, uncovering new error trajectories," while the reported headline numbers are success rates on the R2R-CE and RxR-CE benchmarks. The manuscript gives no guarantee that validation/test splits were never used in any flywheel iteration, nor does it specify the exact evaluation splits, episode sets, or comparison protocol. Without this information, the "new state-of-the-art" claim is open to training-set contamination. Evaluation-split hygiene must be documented precisely, including which splits are used for data generation and which for final evaluation.
  3. [Abstract] The self-correction loop relies on automatically generated labels as the training signal. The abstract provides no description of how deviations are identified or how perception/action corrections are generated, and no accuracy, precision, recall, or human-verification rate for these pseudo-labels. If the deviation detector has low precision, the flywheel can reinforce the model's own errors rather than correct them. This is the load-bearing premise of the entire paradigm and needs quantitative support, for example precision/recall of the deviation detector, label agreement studies, and ablations that compare flywheel training against static training on the same data.
  4. [Abstract] The reported 65.1% and 69.3% success rates are single-point numbers with no variance, number of seeds, or comparison table. Navigation benchmark results are known to vary with random seeds and episode selection; a single-point comparison to prior work is insufficient to support a state-of-the-art claim. Any revised submission would need seed-level statistics, confidence intervals, and a full baseline table on the standard splits.
minor comments (2)
  1. [Abstract] In the final sentence, "\method" is an unexpanded LaTeX macro; it should read "CorrectNav."
  2. [Full text] The arXiv metadata and the PDF content are inconsistent: the header identifies a different paper. Before any further review, the correct PDF must be uploaded; this is not a stylistic issue but a submission-integrity problem.

Circularity Check

0 steps flagged

No circularity detectable from abstract; full text mismatch prevents deeper check.

full rationale

The only CorrectNav content available is the abstract. The abstract describes a self-training flywheel in which the model's error trajectories on the training set are used to generate self-correction data for continued training, followed by re-evaluation on the training set to find new errors. This is a data-generation and training procedure, not a derivation or a prediction reduced to a fit. The central empirical claim—new state-of-the-art success rates of 65.1% on R2R-CE and 69.3% on RxR-CE—is evaluated on external benchmarks, not on the self-generated training data, so the flywheel loop is not the evaluation target. No equations, no fitted parameters, no self-citations, and no uniqueness theorems are present in the abstract that could constitute a circular step. The supplied full text is a different paper (arXiv:2508.10421 on Chinese idiom translation), so the CorrectNav derivation chain cannot be inspected further. Without a specific reduction or a fitted input renamed as a prediction, no circularity is established. The appropriate finding is no significant circularity (score 0).

Axiom & Free-Parameter Ledger

2 free parameters · 3 axioms · 0 invented entities

Abstract-only ledger. The flywheel paradigm contributes a retraining loop but assumes without visible evidence that automatically identified deviations and generated correction labels are trustworthy (ad hoc to the paper), that monocular RGB input suffices (domain assumption), and that training-split errors transfer to evaluation (domain assumption). Two hyperparameters (flywheel iteration count, deviation criterion) are unspecified. The paper introduces no new physical entities; CorrectNav and the Self-correction Flywheel are a model and a training scheme, and their falsifiable handle is the claimed benchmark and robot performance, which this review could not check.

free parameters (2)
  • number of flywheel iterations
    Not stated in the abstract; the reported success rates presumably depend on how many self-correction cycles were run, and this stopping point is a free choice.
  • deviation identification threshold or criterion
    The abstract says deviations in error trajectories are identified, but gives no criterion; the volume and quality of generated correction data, and therefore the claimed gains, depend on it.
axioms (3)
  • ad hoc to paper Error trajectories on the training set contain a learnable corrective signal that can be extracted automatically and without human verification
    This is the engine of the flywheel. The abstract asserts that deviations can be identified and that self-correction data are automatically generated, but provides no evidence that these pseudo-labels are accurate enough to train on.
  • domain assumption A single monocular RGB camera provides sufficient perception for the navigation task
    CorrectNav is described in the abstract as monocular RGB-based; the entire pipeline rests on this sensor choice.
  • domain assumption Training-split error distributions from R2R-CE and RxR-CE transfer to evaluation-time navigation errors
    The flywheel improves the model by re-measuring it on the training set; the abstract does not demonstrate that training-set deviations are the errors that matter on the benchmark evaluation splits or on real robots.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of CorrectNav: Self-Correction Flywheel Empowers Vision-Language-Action Navigation Model." pith.science (2026). https://pith.science/paper/JAKPY3YF

@misc{pith2026250810416,
  author       = {Pith},
  title        = {Pith review of: CorrectNav: Self-Correction Flywheel Empowers Vision-Language-Action Navigation Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JAKPY3YF}},
  note         = {Machine review of arXiv:2508.10416}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Existing vision-and-language navigation models often deviate from the correct trajectory when executing instructions. However, these models lack effective error correction capability, hindering their recovery from errors. To address this challenge, we propose Self-correction Flywheel, a novel post-training paradigm. Instead of considering the model's error trajectories on the training set as a drawback, our paradigm emphasizes their significance as a valuable data source. We have developed a method to identify deviations in these error trajectories and devised innovative techniques to automatically generate self-correction data for perception and action. These self-correction data serve as fuel to power the model's continued training. The brilliance of our paradigm is revealed when we re-evaluate the model on the training set, uncovering new error trajectories. At this time, the self-correction flywheel begins to spin. Through multiple flywheel iterations, we progressively enhance our monocular RGB-based VLA navigation model CorrectNav. Experiments on R2R-CE and RxR-CE benchmarks show CorrectNav achieves new state-of-the-art success rates of 65.1% and 69.3%, surpassing prior best VLA navigation models by 8.2% and 16.4%. Real robot tests in various indoor and outdoor environments demonstrate \method's superior capability of error correction, dynamic obstacle avoidance, and long instruction following.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 11 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. AwareVLN: Reasoning with Self-awareness for Vision-Language Navigation

    cs.RO 2026-05 unverdicted novelty 7.0

    AwareVLN introduces a structural reasoning module and automatic data engine with progress division to equip VLN agents with self-awareness of agent state and task progress, outperforming prior methods on Habitat datasets.

  2. Dual-Anchoring: Addressing State Drift in Vision-Language Navigation

    cs.CV 2026-04 unverdicted novelty 7.0

    Dual-Anchoring Framework mitigates progress drift via structured instruction tokens and memory drift via landmark-centric retrospective prediction, yielding 15.2% success rate gain and 24.7% on long trajectories.

  3. SpikeVLA: Vision-Language-Action Models with Spiking Neural Networks

    cs.RO 2026-06 unverdicted novelty 6.0

    SpikeVLA replaces transformer components in VLA models with spiking vision encoder, multi-modal LLM, and action policy network to reduce energy consumption while maintaining competitive performance on navigation tasks.

  4. Escaping the Diversity Trap in Robotic Manipulation via Anchor-Centric Adaptation

    cs.RO 2026-05 unverdicted novelty 6.0

    Anchor-Centric Adaptation escapes the diversity trap by prioritizing repeated demonstrations at core anchors over broad coverage, yielding higher success rates under fixed data budgets in robotic manipulation.

  5. Dual-Anchoring: Addressing State Drift in Vision-Language Navigation

    cs.CV 2026-04 conditional novelty 6.0

    Privatar partitions VR avatar reconstruction via frequency-domain decomposition, keeping sensitive components local and offloading the rest with distribution-aware minimal perturbation noise, achieving 2.37x throughpu...

  6. HIMM: Human-Inspired Long-Term Memory Modeling for Embodied Exploration and Question Answering

    cs.RO 2026-02 conditional novelty 6.0

    An embodied-agent memory framework that disentangles episodic and semantic memories, retrieves past experiences via visual reasoning, and distills program-style rules achieves new state-of-the-art results on A-EQA and...

  7. AstraNav-World: World Model for Foresight Control and Consistency

    cs.CV 2025-12 unverdicted novelty 6.0

    AstraNav-World unifies diffusion video generation and vision-language action planning in a single bidirectional model that improves trajectory accuracy, success rates, and zero-shot real-world adaptation in embodied n...

  8. Efficient-VLN: A Simple yet Strong Baseline for Efficient Vision-Language Navigation

    cs.CV 2025-12 reject novelty 6.0

    The submitted abstract, the paper body, and the conclusion report incompatible headline results (73.2/75.6, 64.2/67.0, and 62.3/64.5 SR on R2R-CE/RxR-CE), so the claimed state of the art is not internally consistent.

  9. What Limits Vision-and-Language Navigation ?

    cs.RO 2026-05 unverdicted novelty 5.0

    StereoNav reaches new benchmark highs on R2R-CE and RxR-CE and improves real-robot reliability by supplying persistent target-location priors and stereo-derived geometry that stay stable under lighting changes and blur.

  10. Dual-Anchoring: Addressing State Drift in Vision-Language Navigation

    cs.CV 2026-04 unverdicted novelty 5.0

    Dual-Anchoring adds explicit progress tokens and retrospective landmark verification to VLN agents, cutting state drift and lifting success rate 15.2% overall with 24.7% gains on long trajectories.

  11. Nav-R1: Reasoning and Navigation in Embodied Scenes

    cs.RO 2025-09 reject novelty 4.0

    Nav-R1 uses a 110K synthetic CoT dataset, GRPO with three rewards, and a fast-in-slow system to set new SOTA on R2R-CE, RxR-CE, and HM3D-OVON.

Reference graph

Works this paper leans on

42 extracted references · 26 canonical work pages · cited by 9 Pith papers

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    An, D.; Qi, Y.; Li, Y.; Huang, Y.; Wang, L.; Tan, T.; and Shao, J. 2023. Bevbert: Multimodal map pre-training for language-guided navigation. In ICCV

    work page 2023

  4. [4]

    An, D.; Wang, H.; Wang, W.; Wang, Z.; Huang, Y.; He, K.; and Wang, L. 2024. Etpnav: Evolving topological planning for vision-language navigation in continuous environments. IEEE TPAMI

    work page 2024

  5. [5]

    An, D.; Wang, Z.; Li, Y.; Wang, Y.; Hong, Y.; Huang, Y.; Wang, L.; and Shao, J. 2022. 1st place solutions for rxr-habitat vision-and-language navigation competition. In CVPRW

    work page 2022

  6. [6]

    Anderson, P.; Wu, Q.; Teney, D.; Bruce, J.; Johnson, M.; S \"u nderhauf, N.; Reid, I.; Gould, S.; and van den Hengel, A. 2018. Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    work page 2018

  7. [7]

    Chang, A.; Dai, A.; Funkhouser, T.; Halber, M.; Niessner, M.; Savva, M.; Song, S.; Zeng, A.; and Zhang, Y. 2017. Matterport3D: Learning from RGB-D Data in Indoor Environments. International Conference on 3D Vision (3DV)

    2017

  8. [8]

    Chen, J.; Lin, B.; Liu, X.; Liang, X.; and Wong, K.-Y. K. 2024. Affordances-Oriented Planning using Foundation Models for Continuous Vision-Language Navigation. arXiv preprint

    work page 2024

  9. [9]

    K.; Chuang, J.; Vázquez, M.; and Savarese, S

    Chen, K.; Chen, J. K.; Chuang, J.; Vázquez, M.; and Savarese, S. 2021. Topological planning with transformers for vision-and-language navigation. In CVPR

    work page 2021

  10. [10]

    Chen, P.; Ji, D.; Lin, K.; Zeng, R.; Li, T.; Tan, M.; and Gan, C. 2022. Weakly-supervised multi-granularity map learning for vision-and-language navigation. In NeurIPS

    work page 2022

  11. [11]

    Cheng, A.-C.; Ji, Y.; Yang, Z.; Gongye, Z.; Zou, X.; Kautz, J.; B y k, E.; Yin, H.; Liu, S.; and Wang, X. 2024. Navila: Legged robot vision-language-action model for navigation. arXiv preprint arXiv:2412.04453

    Pith/arXiv arXiv 2024

  12. [12]

    R.; Tian, S.; Yuan, W.; Krishna, R.; Fox, D.; Mandlekar, A.; and Guo, Y

    Duan, J.; Pumacay, W.; Kumar, N.; Wang, Y. R.; Tian, S.; Yuan, W.; Krishna, R.; Fox, D.; Mandlekar, A.; and Guo, Y. 2024. AHA: A vision-language-model for detecting and reasoning over failures in robotic manipulation. arXiv preprint arXiv:2410.00371

    Pith/arXiv arXiv 2024

  13. [13]

    Georgakis, G.; Schmeckpeper, K.; Wanchoo, K.; Dan, S.; Miltsakaki, E.; Roth, D.; and Daniilidis, K. 2022. Cross-modal map learning for vision and language navigation. In CVPR

    work page 2022

  14. [14]

    Ha, H.; Florence, P.; and Song, S. 2023. Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition. arXiv:2307.14535

    Pith/arXiv arXiv 2023

  15. [15]

    Hong, Y.; Wang, Z.; Wu, Q.; and Gould, S. 2022. Bridging the gap between learning in discrete and continuous environments for vision-and-language navigation. In CVPR

    work page 2022

  16. [16]

    Hong, Y.; Zhou, Y.; Zhang, R.; Dernoncourt, F.; Bui, T.; Gould, S.; and Tan, H. 2023. Learning navigational visual representations with semantic map supervision. In ICCV

    work page 2023

  17. [17]

    Ilharco, G.; Jain, V.; Ku, A.; Ie, E.; and Baldridge, J. 2019. General Evaluation for Instruction Conditioned Navigation using Dynamic Time Warping. arXiv:1907.05446

    Pith/arXiv arXiv 2019

  18. [18]

    Krantz, J.; Gokaslan, A.; Batra, D.; Lee, S.; and Maksymets, O. 2021. Waypoint models for instruction-guided navigation in continuous environments. In CVPR

    work page 2021

  19. [19]

    Krantz, J.; and Lee, S. 2022. Sim-2-sim transfer for vision-and-language navigation in continuous environments. In ECCV

    work page 2022

  20. [20]

    Krantz, J.; Wijmans, E.; Majundar, A.; Batra, D.; and Lee, S. 2020. Beyond the Nav-Graph: Vision and Language Navigation in Continuous Environments. In ECCV

    work page 2020

  21. [21]

    Ku, A.; Anderson, P.; Patel, R.; Ie, E.; and Baldridge, J. 2020. Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense Spatiotemporal Grounding. In EMNLP

    work page 2020

  22. [22]

    L.; Wei, Z.; Han, M.; Xu, R.; Niu, M.; Han, J.; Lin, L.; Lu, C.; and Liang, X

    Lin, B.; Nie, Y.; Zai, K. L.; Wei, Z.; Han, M.; Xu, R.; Niu, M.; Han, J.; Lin, L.; Lu, C.; and Liang, X. 2025. EvolveNav: Self-Improving Embodied Reasoning for LLM-Based Vision-Language Navigation. arXiv:2506.01551

    arXiv 2025

  23. [23]

    Liu, H.; Li, C.; Li, Y.; and Lee, Y. J. 2024. Improved baselines with visual instruction tuning. In CVPR

    work page 2024

  24. [24]

    Liu, Z.; Bahety, A.; and Song, S. 2023. REFLECT: Summarizing Robot Experiences for Failure Explanation and Correction. arXiv preprint arXiv:2306.15724

    Pith/arXiv arXiv 2023

  25. [25]

    Long, Y.; Cai, W.; Wang, H.; Zhan, G.; and Dong, H. 2024. InstructNav: Zero-shot System for Generic Instruction Navigation in Unexplored Environment. arXiv:2406.04882

    Pith/arXiv arXiv 2024

  26. [26]

    J.; Sodhani, S.; Jayaraman, D.; Bastani, O.; Kumar, V.; and Zhang, A

    Ma, Y. J.; Sodhani, S.; Jayaraman, D.; Bastani, O.; Kumar, V.; and Zhang, A. 2023. VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training. arXiv:2210.00030

    Pith/arXiv arXiv 2023

  27. [27]

    D.; Partsey, R.; Yang, J.; Desai, R.; Clegg, A

    Puig, X.; Undersander, E.; Szot, A.; Cote, M. D.; Partsey, R.; Yang, J.; Desai, R.; Clegg, A. W.; Hlavac, M.; Min, T.; Gervet, T.; Vondruš, V.; Berges, V.-P.; Turner, J.; Maksymets, O.; Kira, Z.; Kalakrishnan, M.; Malik, J.; Chaplot, D. S.; Jain, U.; Batra, D.; Rai, A.; and Mottaghi, R. 2023. Habitat 3.0: A Co-Habitat for Humans, Avatars and Robots

    work page 2023

  28. [28]

    Raychaudhuri, S.; Wani, S.; Patel, S.; Jain, U.; and Chang, A. 2021. Language-Aligned Waypoint (LAW) Supervision for Vision-and-Language Navigation in Continuous Environments. In EMNLP

    work page 2021

  29. [29]

    Shi, X.; Li, Z.; Lyu, W.; Xia, J.; Dayoub, F.; Qiao, Y.; and Wu, Q. 2025. SmartWay: Enhanced Waypoint Prediction and Backtracking for Zero-Shot Vision-and-Language Navigation. arXiv:2503.10069

    Pith/arXiv arXiv 2025

  30. [30]

    Wang, H.; Liang, W.; Van Gool, L.; and Wang, W. 2023 a . Dreamwalker: Mental planning for continuous vision-language navigation. In ICCV

    work page 2023

  31. [31]

    Wang, Z.; Li, J.; Hong, Y.; Wang, Y.; Wu, Q.; Bansal, M.; Gould, S.; Tan, H.; and Qiao, Y. 2023 b . Scaling data generation in vision-and-language navigation. In ICCV

    work page 2023

  32. [32]

    Wang, Z.; Li, X.; Yang, J.; Liu, Y.; Hu, J.; Jiang, M.; and Jiang, S. 2024. Lookahead Exploration with Neural Radiance Representation for Continuous Vision-Language Navigation. In CVPR

    work page 2024

  33. [33]

    Wang, Z.; Li, X.; Yang, J.; Liu, Y.; and Jiang, S. 2023 c . Gridmm: Grid memory map for vision-and-language navigation. In ICCV

    work page 2023

  34. [34]

    Wei, M.; Wan, C.; Yu, X.; Wang, T.; Yang, Y.; Mao, X.; Zhu, C.; Cai, W.; Wang, H.; Chen, Y.; Liu, X.; and Pang, J. 2025 a . StreamVLN: Streaming Vision-and-Language Navigation via SlowFast Context Modeling. arXiv:2507.05240

    Pith/arXiv arXiv 2025

  35. [35]

    Wei, Z.; Lin, B.; Nie, Y.; Chen, J.; Ma, S.; Xu, H.; and Liang, X. 2025 b . Unseen from Seen: Rewriting Observation-Instruction Using Foundation Models for Augmenting Vision-Language Navigation. arXiv:2503.18065

    work page arXiv 2025

  36. [36]

    Xiao, J.; Shang, X.; Yao, A.; and Chua, T.-S. 2021. NExT-QA: Next Phase of Question-Answering to Explaining Temporal Actions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 9777--9786

    work page 2021

  37. [37]

    Yang, A.; Yang, B.; Hui, B.; Zheng, B.; Yu, B.; Zhou, C.; Li, C.; Li, C.; Liu, D.; Huang, F.; Dong, G.; Wei, H.; Lin, H.; Tang, J.; Wang, J.; Yang, J.; Tu, J.; Zhang, J.; Ma, J.; Yang, J.; Xu, J.; Zhou, J.; Bai, J.; He, J.; Lin, J.; Dang, K.; Lu, K.; Chen, K.; Yang, K.; Li, M.; Xue, M.; Ni, N.; Zhang, P.; Wang, P.; Peng, R.; Men, R.; Gao, R.; Lin, R.; Wan...

    Pith/arXiv arXiv 2024

  38. [38]

    Yu, Z.; Xu, D.; Yu, J.; Yu, T.; Zhao, Z.; Zhuang, Y.; and Tao, D. 2019. ActivityNet-QA: A Dataset for Understanding Complex Web Videos via Question Answering. In AAAI, 9127--9134

    work page 2019

  39. [39]

    Zhai, X.; Mustafa, B.; Kolesnikov, A.; and Beyer, L. 2023. Sigmoid loss for language image pre-training. In ICCV

    work page 2023

  40. [40]

    Zhang, J.; Wang, K.; Wang, S.; Li, M.; Liu, H.; Wei, S.; Wang, Z.; Zhang, Z.; and Wang, H. 2025. Uni-NaVid: A Video-based Vision-Language-Action Model for Unifying Embodied Navigation Tasks. arXiv:2412.06224

    Pith/arXiv arXiv 2025

  41. [41]

    Zhang, J.; Wang, K.; Xu, R.; Zhou, G.; Hong, Y.; Fang, X.; Wu, Q.; Zhang, Z.; and He, W. 2024 a . NaVid: Video-based VLM Plans the Next Step for Vision-and-Language Navigation. In RSS

    work page 2024

  42. [42]

    Zhang, Y.; Wu, J.; Li, W.; Li, B.; Ma, Z.; Liu, Z.; and Li, C. 2024 b . Video Instruction Tuning With Synthetic Data. arXiv:2410.02713

    Pith/arXiv arXiv 2024

This paper was first reviewed by deepseek-v4-flash on August 5, 2026.