Pith. sign in

REVIEW 3 major objections 5 minor 80 references

Goal-oriented Navigation Instruction Generation with Tour Video Priors

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper attempts to show that ego-centric tour videos, an initial observation, and a goal description are enough to generate executable navigation instructions, without maps or graphs.

desk verdict VideoNIG is a genuinely new benchmark task with a careful construction pipeline, but the headline MPO gains on Choice Evaluation are partly by construction because training and evaluation share the same distractor taxonomy. read the letter →

arxiv 2608.08596 v1 pith:UN7HXFPU submitted 2026-08-09 cs.CV

Fangdi Li , Juncheng Liao , Changxu Cheng , Jiazhi Wang , Senda Chen , Tao Wang , Wuyue Zhao This is my paper · ORCID
classification cs.CV
keywords navigationinstructiongenerationvision-and-languagemultimodallargelanguagemodelscurriculumlearningego-centrictourvideospatialreasoningbenchmarkHabitatsimulator
verification ladder T0 review T1 audit T2 compute T3 formal

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 introduces VideoNIG, a goal-oriented navigation instruction generation task where a model must produce step-by-step instructions from an ego-centric tour video, a single initial observation, and a textual or image goal, with no map or graph available. It instantiates the task in a controlled Habitat-based benchmark with roughly 60K tour videos and 37K multimodal prompts, and evaluates outputs through text similarity, distractor-based choice tests, and downstream navigation execution by a VLN agent. The central claim is that this video-grounded formulation is viable but challenging, and that current multimodal LLMs struggle with it, particularly when the tour does not simply follow the optimal route. The paper further argues that a two-stage curriculum, first teaching action-view alignment and then increasing trajectory complexity, substantially improves instruction quality and executability.

What carries the argument

The central mechanism is the two-stage Curriculum Learning framework. The first stage, Action Warmup, uses paired tasks, action-from-view and view-from-action, to align the model's spatial representations with discrete ego-centric motions before it encounters long narratives. The second stage, Complexity Progression, orders training data from Gold Route and Tyro Tour videos through Curiosity Tour to the more deviating Explorer Tour videos, so the model gradually handles longer and more exploratory trajectories. The benchmark's tour construction, which guarantees each tour covers the Gold Route's viewpoints while adding start/goal extensions, look-arounds, and neighborhood visits, is what makes the task well-posed.

What would settle it

Take an Explorer Tour sample, remove the frames that cover one middle corridor of the Gold Route, and ask the trained model to produce instructions for that corridor; if the model still names objects in the unseen segment, it is relying on memorized layout priors rather than video grounding, and if it fails, the benchmark's route-coverage assumption is load-bearing for its results.

Watch

Extended reading notes

Core claim

VideoNIG defines a task in which the environmental prior is a continuous ego-centric video rather than a graph, map, or landmark set, and it shows that a multimodal LLM can be trained to plan the optimal subpath inside that video and verbalize it as navigation instructions. The benchmark covers three tour complexities, from simple endpoint extensions to tours with substantial detours, and the paper's experiments show that performance degrades predictably as tour complexity rises. The proposed two-stage curriculum, Action Warmup followed by Complexity Progression, is demonstrated to improve spatial grounding on choice-based tests and to raise navigation success when the generated instructions are executed by a VLN agent; for example, on R2R Gold Route the MPO-trained model raises Success Rate from 40.3 to 56.1 in the System2 setting. The paper also claims that choice-based spatial consistency metrics correlate more strongly with downstream navigation success than traditional text similarity metrics.

Load-bearing premise

Every tour video is constructed so that it visually covers the full optimal route's viewpoints, which means the correct instruction can in principle be read out of the video; if a tour missed a necessary segment of the optimal path, the task could not be solved from the video alone.

Editorial extensions

If this is right

  • If the task formulation is correct, goal-oriented navigation instruction generation can be decoupled from explicit environmental representations, since richly recorded videos can serve as the spatial prior.
  • Tour-video complexity, especially route deviation and exploratory behavior, is a scaling bottleneck for current MLLMs, so progress on VideoNIG should track with better long-horizon video grounding.
  • Choice-based spatial consistency tests, such as the proposed Multiple-choice and Orthogonal-choice evaluations, are better predictors of instruction executability than Rouge-L and SPICE, making them useful diagnostic tools for future instruction-generation research.
  • Training that first reinforces local action-view alignment and then increases trajectory complexity yields stable gains across text similarity, choice accuracy, and downstream VLN success, suggesting that curriculum order matters for spatial reasoning tasks.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the central claim holds, a practical device could pre-record a single walk-through of an indoor space and later answer arbitrary goal queries with turn-by-turn instructions, without ever building a semantic map.
  • The choice-based metrics may transfer to other instruction-generation domains where an output is judged by whether a downstream policy can execute it, not by surface similarity to a reference.
  • The route-coverage assumption is the key boundary: extending VideoNIG to naturally collected tour videos that may miss parts of the optimal path would likely expose hallucination and would be a direct stress test of the task's grounding claim.
  • One testable extension would be to replace the tour-type curriculum with a continuous difficulty measure, such as circuitousness or detour ratio, which might yield smoother learning and better generalization than the discrete Gold/Tyro/Curiosity/Explorer ordering.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces VideoNIG, a task in which a model must generate step-by-step navigation instructions from an ego-centric tour video, an initial observation, and a textual or visual goal, without using maps or graphs. The authors instantiate this task in a Habitat simulator benchmark built from R2R-CE and RxR-CE, with four video types (Gold Route, Tyro Tour, Curiosity Tour, Explorer Tour) of increasing complexity, together with multimodal start-goal descriptions. They propose a two-stage Curriculum Learning framework consisting of Action Warmup (action-view alignment) followed by Complexity Progression (training on progressively harder tours), and they train Qwen3-VL-8B with SFT, GRPO, and MPO. Evaluation uses Rouge-L/SPICE text similarity, Multiple-choice and Orthogonal-choice spatial-consistency tests, and downstream navigation execution with InternVLA-N1. The central empirical claims are that existing MLLMs struggle on VideoNIG and that the proposed curriculum training improves instruction quality across the diagnostic metrics.

Significance. If the results hold, VideoNIG would be a useful controlled benchmark for video-grounded spatial reasoning in navigation, and the curriculum training pipeline would be a reasonable reference method. Strengths include the large-scale constructed benchmark (about 60K tour videos), the graded difficulty design, the combination of three evaluation perspectives, and the explicit acknowledgment of simulator-only settings and of the route-coverage assumption. The navigation execution results provide at least partially independent evidence that instructions produced by the MPO-trained model are more executable, especially on Gold Route. However, the two text-based evidence channels are partially coupled to the training objectives (GRPO rewards the same text metrics; MPO negatives share the same distractor taxonomy as the choice tests), so the magnitude of the claimed improvements over baselines is not yet established in a fully unbiased way.

major comments (3)
  1. [Sec. 5.1, App. C.2 vs Sec. 3.4] The MPO training and the Choice Evaluation share the same five distractor types: direction, entity, deletion, addition, change, and swap. Appendix C.2 states that MPO's rejected instructions are generated by qwen3-vl-plus using these five types, and Sec. 3.4 defines Multiple-choice and Orthogonal-choice with exactly the same categories. Tab. 2 then shows MPO-lora improving average Choice-Evaluation accuracy to roughly 60.6/62.0 from SFT's 49.8/53.1. Because Sec. 5.2 explicitly relies on Choice Evaluation as the diagnostic that 'better reflects practical navigation performance' and as the main evidence that the curriculum helps, these gains may reflect overfitting to the evaluation taxonomy rather than general video-grounded spatial reasoning. Please add a held-out evaluation with distractor types not used in MPO training (or generated by a different procedure) and report per-type accuracies.
  2. [Eq. (2), Tab. 2] The GRPO reward in Eq. (2) includes fSPICE and fRouge-L, and Tab. 2 reports SPICE and Rouge-L as the text-similarity evidence of improvement. Since the optimization directly maximizes these metrics, the higher GRPO text-similarity scores (e.g., average Rouge-L 0.254 vs SFT 0.233) are at least partly by construction. This is acknowledged in Sec. 5.2, but the paper still summarizes the results as 'significantly improves instruction quality across complementary diagnostic metrics' (Abstract). Please either present text-similarity scores as training objectives rather than independent evaluation evidence, or ablate the similarity terms from the reward and show that improvements persist.
  3. [Tab. 3] The navigation execution results that are cleanly independent of the training/evaluation overlaps support MPO mainly on Gold Route (e.g., R2R SR 56.1 vs 40.3 for System2 baseline), while improvements on Tyro, Curiosity, and Explorer tours are small (e.g., 33.2 vs 31.0 on Tyro, 21.2 vs 21.0 on Curiosity, and 20.4 vs 18.7 on Explorer). The paper does not report variance estimates or significance tests, and it is unclear how many episodes underlie each number. Please provide standard errors or confidence intervals and state the number of evaluation episodes.
minor comments (5)
  1. [Sec. 3.2] The text says 'We render four types of videos' but the reader sees three tour types plus the Gold Route; consider clarifying 'four video types, including the Gold Route and three tour variants'.
  2. [Tab. 2] Table 2 is extremely wide and difficult to parse because text-similarity and choice-evaluation numbers are interleaved; splitting it into separate tables for text metrics and choice metrics (or reordering columns) would improve readability.
  3. [App. D.1] Proprietary models are evaluated on only 100 randomly sampled instances per configuration; the choice-evaluation results for these models in Tab. 7 are therefore noisy. Please describe the sampling procedure and ideally report confidence intervals or a larger sample.
  4. [Sec. 3.3] The term 'circuitousness' is nonstandard; consider using 'detour ratio' or defining it more explicitly as the route-to-Euclidean distance ratio.
  5. [Sec. 5.2] The claim that 'the GRPO-trained model achieves higher Rouge-L and SPICE scores' is stated relative to SFT, but on some rows MPO has higher Rouge-L (e.g., Gold Route R2R image goal); the sentence should be reworded to avoid ambiguity about which comparison is meant.

Circularity Check

2 steps flagged · score 6.0 of 10

Headline diagnostic gains are partly by construction: GRPO rewards the exact text-similarity metrics reported in Tab. 2, and MPO trains on the same distractor taxonomy used in Choice Evaluation; navigation-execution results remain independent evidence.

  1. fitted input called prediction [Sec. C.1, Eq. (2); Sec. 5.2/Tab. 2; Sec. D.1]
    "The main reward component is defined as: Rmain=αfSPICE(Ipred,Igt)+βfRouge-L(Ipred,Igt)+γfquality(Ipred)+δftraj(Vgt,Ipred)... GRPO-trained models generally achieve the highest text-similarity scores, reflecting superior alignment with the reference instruction’s surface-level phrasing. This is expected, as the GRPO reward function explicitly optimizes these metrics during training."

    The GRPO objective maximizes Eq. (2), which contains fSPICE and fRouge-L with positive weights, so training directly pushes up those two scores on the training distribution. Table 2 then reports exactly these same two metrics as evidence that the trained model improves instruction quality. The Rouge-L/SPICE gains are therefore enforced by the reward definition rather than independently measured, as the paper itself concedes in Sec. D.1. This does not invalidate the whole paper: the quality, trajectory-consistency, and length terms are separate, and downstream navigation execution is not part of the reward, so the circularity is partial.

  2. fitted input called prediction [Sec. 5.1 and Sec. C.2 (MPO training); Sec. 3.4, Fig. 4 (Choice Evaluation)]
    "MPO employs preference-based learning using negative samples constructed via our evaluation protocol (see Fig. 4), where one negative sample is randomly selected per instance. ... Specifically, we design five types of negative samples as rejected instructions (Fig. 4)."

    Section 3.4 defines Multiple-choice and Orthogonal-choice using the five distractor types shown in Fig. 4 (direction, entity, deletion, addition, change, swap), and Sec. C.2 constructs the MPO rejected-instruction pool from the same five types via qwen3-vl-plus. The model is therefore trained to reject the same distractor taxonomy on which its headline Choice Evaluation accuracy is measured. The large MPO-vs-SFT gap in Tab. 2 (e.g., MPO-lora averages around 60.6/62.0 vs 49.8/53.1) may thus reflect training/evaluation overlap rather than improved general video-grounded spatial reasoning. Orthogonal-choice still requires choosing among unseen instances, so the reduction is not total, but the diagnostic is no longer independent of the training signal.

full rationale

The paper contains two partial circularities. First, Eq. (2) places fSPICE and fRouge-L in the GRPO reward, and Tab. 2 reports exactly those same metrics as evidence of improvement; the authors explicitly state that the gain is expected because the reward optimizes those metrics. Second, MPO's rejected instructions are generated with the same five distractor types that define the Multiple-choice/Orthogonal-choice evaluation, so the strong Choice-Evaluation advantage of MPO over SFT is partly attributable to training on the evaluation taxonomy rather than to independently measured spatial reasoning. Neither reduction is complete: the trajectory-consistency and LLM-judge reward terms are not among the reported evaluation metrics, the Orthogonal-choice test still uses unseen instances, and the downstream VLN execution results in Tab. 3 (e.g., R2R SR rising from 40.3 to 56.1 for MPO on Gold Route) are not encoded in any training objective. There is no load-bearing self-citation chain or uniqueness-theorem argument; the authors' prior work is used for training recipes and the downstream VLN follower, not to justify the task's central claim. Overall, two headline diagnostics are partly by construction, but independent execution evidence supports the core executability claim, so a score of 6 rather than 8 or 10 is appropriate.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The empirical claims rest on route-covering synthetic tours, filtered human annotations, and unvalidated MLLM judges. Training introduces hand-chosen reward weights and generation distributions. These are stated in the paper, but they bound the generality of the conclusions.

free parameters (5)
  • GRPO reward weights (alpha, beta, gamma, delta) = 0.1, 0.1, 0.2, 0.6
    Hand-chosen in Eq. (2); no sensitivity analysis reported. The dominant weight 0.6 is an MLLM trajectory-consistency score.
  • Tour offset distributions = Tyro: Uniform(1,4); Curiosity: Normal(3.6,1.2); Explorer: Normal(7.5,2.5)
    Appendix A defines difficulty levels by these hand-specified distributions; different offsets would change the benchmark difficulty gradient.
  • Exploration trigger probability P = 0.02 in offset regions, 0.07 in Gold Route region
    Appendix A.5 sets how often look-around and neighborhood visits occur; this controls tour diversity and task difficulty.
  • Gold Route optimality threshold nDTW = 0.8
    Sec. 3.2 uses nDTW below 0.8 to discard non-optimal routes; changing the threshold changes which samples remain in the benchmark.
  • Length penalty thresholds in GRPO reward = r<3 gives 0; 3<=r<=5 gives -(r-3)/2; r>5 gives -1
    Eq. (3) is a hand-designed penalty on generation length relative to ground truth; it affects the text length of GRPO outputs.
assumptions (4)
  • domain assumption Tour videos contain the optimal Gold Route viewpoints by construction.
    Sec. 3.2 states that tours ensure coverage of the optimal path's viewpoints. The task is therefore route-covering video selection, not planning from arbitrary priors. The paper limits claims to simulator-based indoor settings.
  • domain assumption Human annotations in R2R-CE and RxR-CE are valid ground truth after the paper's filtering.
    The benchmark uses existing annotations as ground truth and computes Rouge-L and SPICE against them. The paper filters roughly one-third of instructions for missing turn-around actions, indicating the raw annotations have systematic flaws.
  • domain assumption The MLLM trajectory-consistency judge and LLM quality judge are reliable reward signals.
    Eq. (2) gives trajectory consistency a weight of 0.6 using an MLLM judge, with no reported validation of judge accuracy against human ratings. A biased judge would bias the reward and therefore the trained policy.
  • domain assumption Choice-based distractors capture the spatial errors that matter for navigation.
    The paper argues choice evaluation better reflects navigation performance, but this is supported only by correlation with one VLN agent's success rate, not by a causal or exhaustive analysis of instruction error modes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Goal-oriented Navigation Instruction Generation with Tour Video Priors." pith.science (2026). https://pith.science/paper/UN7HXFPU

@misc{pith2026260808596,
  author       = {Pith},
  title        = {Pith review of: Goal-oriented Navigation Instruction Generation with Tour Video Priors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UN7HXFPU}},
  note         = {Machine review of arXiv:2608.08596}
}
read the original abstract

Navigation Instruction Generation (NIG) aims to produce step-by-step natural language instructions for navigation guidance. Existing studies primarily treat NIG as an auxiliary task for vision-andlanguage navigation (VLN), focusing on data augmentation or multi-task learning. However, generating navigation instructions from compact environmental priors requires meticulous spatial reasoning, especially when the target route does not simply follow the demonstrated tour, and remains challenging for current multimodal models. In this work, we introduce VideoNIG, a goal-oriented video-grounded NIG task that generates navigation instructions from ego-centric tour videos, an initial observation, and a textual or visual goal, without relying on intermediate representations such as graphs and maps. We instantiate VideoNIG in a controlled simulator benchmark with 60K tour videos across continuous indoor environments and 37K multimodal prompts with progressive difficulty levels. We further introduce a diagnostic evaluation protocol that combines text similarity, choice-based spatial consistency tests, and downstream navigation execution. To address this task, we propose a two-stage Curriculum Learning framework that decomposes the learning into foundational motion perception and long-horizon navigation reasoning. Specifically, we first employ Action Warmup for spatial action-view alignment, followed by Complexity Progression using trajectories with increasing exploratory difficulty. Extensive experiments show that existing MLLMs struggle with VideoNIG, while our approach significantly improves instruction quality across complementary diagnostic metrics. Finally, integrating VideoNIG-generated instructions with a VLN agent demonstrates the executability of this task formulation for end-to-end navigation.

Figures

Figures reproduced from arXiv: 2608.08596 by the authors.

Figure 1
Figure 1. The relationship between NIG and VLN. (a) The input to a VLN agent is typi￾cally fine-grained textual guidance, which demands extensive reasoning from users and is therefore not user-friendly. (b) Most previous works formulate NIG as a trajectory￾captioning task, optionally with auxiliary heuristic environmental information. (c) Our VideoNIG agent requires only simple goal descriptions from users, together with sing… view at source ↗
Figure 2
Figure 2. Benchmark construction pipeline for VideoNIG. (a) The left panel shows the observation at the arrow position in the right panel. The orange and red points are the initial and goal positions of the green Gold Route, respectively, and the blue trajectory is an Explorer Tour. (c) Outliers are filtered via three different methods. Blue and green points remain after filtering. (e) The initial and goal descriptions (text … view at source ↗
Figure 3
Figure 3. Data distributions in VideoNIG. (a) and (b) compare the distance and cir￾cuitousness of Gold Route across the filtered R2R and RxR samples. (c)-(h) compare the distributional differences among different tours. 3.2 Benchmark Instantiation We instantiate VideoNIG based on the classical R2R-CE and RxR-CE [4,29,30] in continuous environments through the Habitat simulator [43], using their high￾quality human-annotated na… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Multiple-choice: tests spatio-temporal consistency via diverse distractors. Orthogonal-choice: isolates direction and entity changes to mitigate textual shortcuts. extensions provide broader and more redundant environmental scenes for opti￾mal route planning. Note that…
Figure 5
Figure 5. Figure 5: Correlation between Rouge-L, SPICE, Multiple-choice, Orthogonal-choice, and VLN Success Rate. and Mixed Preference Optimization (MPO) [54]. Both are trained with the proposed Curriculum Learning schedule. For GRPO, the reward function is a weighted combination of four …
Figure 6
Figure 6. Figure 6: Correlation of VSI-Bench, Route Planning, and VideoNIG Performance 5.2 Results on VideoNIG Main Results and Comparative Analysis. We evaluate InternVL and Qwen series models across four metrics: Rouge-L, SPICE, Multiple-choice, and Orthogonal-choice in Tab. 2. Results …
Figure 7
Figure 7. Figure 7: Qualitative analysis of a VideoNIG image-goal case. with SPICE and our diagnostic Orthogonal-choice most aligned with general spatial reasoning. Takeaway: VideoNIG performance strongly tracks visual-spatial reasoning, highlighting it as a key ability for navigation ins…
Figure 8
Figure 8. Figure 8: The prompt templates support goal specification via either image-based goal observations or textual descriptions in the VideoNIG task. trained models also excel, indicating a reduced reliance on superficial textual patterns and an improved focus on core directional and…
Figure 9
Figure 9. Figure 9: Qualitative comparison of navigation instructions generated by proprietary models, open-source models, and our fine-tuned models. phrasing and syntactic patterns. Consequently, even when model-generated in￾structions are spatially accurate, subtle stylistic departures …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

80 extracted references · 42 canonical work pages

  1. [1]

    arXiv preprint arXiv:2212.04385 Goal-oriented Navigation Instruction Generation with Tour Video Priors 15 (2022) 4

    An, D., Qi, Y., Li, Y., Huang, Y., Wang, L., Tan, T., Shao, J.: Bevbert: Multimodal map pre-training for language-guided navigation. arXiv preprint arXiv:2212.04385 Goal-oriented Navigation Instruction Generation with Tour Video Priors 15 (2022) 4

    arXiv 2022

  2. [2]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (2024) 3, 4

    An, D., Wang, H., Wang, W., Wang, Z., Huang, Y., He, K., Wang, L.: Etpnav: Evolving topological planning for vision-language navigation in continuous envi- ronments. IEEE Transactions on Pattern Analysis and Machine Intelligence (2024) 3, 4

    work page 2024

  3. [3]

    In: European conference on computer vision

    Anderson, P., Fernando, B., Johnson, M., Gould, S.: Spice: Semantic propositional image caption evaluation. In: European conference on computer vision. pp. 382–

  4. [4]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Anderson, P., Wu, Q., Teney, D., Bruce, J., Johnson, M., Sünderhauf, N., Reid, I., Gould, S., Van Den Hengel, A.: Vision-and-language navigation: Interpreting visually-grounded navigation instructions in real environments. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 3674–3683 (2018) 1, 4, 6

    work page 2018

  5. [5]

    In: 2025 IEEE International Conference on Robotics and Automation (ICRA)

    Anwar, A., Welsh, J., Biswas, J., Pouya, S., Chang, Y.: Remembr: Building and reasoning over long-horizon spatio-temporal memory for robot navigation. In: 2025 IEEE International Conference on Robotics and Automation (ICRA). pp. 2838–

    2025

  6. [6]

    5-vl technical report

    Bai, S., Chen, K., Liu, X., Wang, J., Ge, W., Song, S., Dang, K., Wang, P., Wang, S., Tang, J., et al.: Qwen2. 5-vl technical report. arXiv preprint arXiv:2502.13923 (2025) 10, 6

    arXiv 2025

  7. [7]

    (2025) 5, 10, 11, 6, 7

    Bai, S., et al.: Qwen3-vl github repo. (2025) 5, 10, 11, 6, 7

    work page 2025

  8. [8]

    In: Pro- ceedings of the 26th annual international conference on machine learning

    Bengio, Y., Louradour, J., Collobert, R., Weston, J.: Curriculum learning. In: Pro- ceedings of the 26th annual international conference on machine learning. pp. 41–48 (2009) 9

    work page 2009

Show all 80 references
  1. [9]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Bigverdi, M., Luo, Z., Hsieh, C.Y., Shen, E., Chen, D., Shapiro, L.G., Krishna, R.: Perception tokens enhance visual reasoning in multimodal language models. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 3836–3845 (2025) 5

    2025

  2. [10]

    arXiv preprint arXiv:2402.11684 (2024) 10

    Chen,G.H.,Chen,S.,Zhang,R.,Chen,J.,Wu,X.,Zhang,Z.,Chen,Z.,Li,J.,Wan, X., Wang, B.: Allava: Harnessing gpt4v-synthesized data for lite vision-language models. arXiv preprint arXiv:2402.11684 (2024) 10

    2024 arXiv

  3. [11]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Chen, S., Guhur, P.L., Tapaswi, M., Schmid, C., Laptev, I.: Think global, act local: Dual-scale graph transformer for vision-and-language navigation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 16537–16547 (2022) 1, 4

    2022

  4. [12]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Chen, Y., Ge, Y., Tang, W., Li, Y., Ge, Y., Ding, M., Shan, Y., Liu, X.: Moto: Latent motion token as the bridging language for learning robot manipulation from videos. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 19752–19763 (2025) 3

    2025

  5. [13]

    arXiv preprint arXiv:2412.04453 (2024) 4

    Cheng,A.C.,Ji,Y.,Yang,Z.,Gongye,Z.,Zou,X.,Kautz,J.,Bıyık,E.,Yin,H.,Liu, S., Wang, X.: Navila: Legged robot vision-language-action model for navigation. arXiv preprint arXiv:2412.04453 (2024) 4

    2024 arXiv

  6. [14]

    arXiv preprint arXiv:2407.07775 (2024) 4

    Chiang, H.T.L., Xu, Z., Fu, Z., Jacob, M.G., Zhang, T., Lee, T.W.E., Yu, W., Schenck, C., Rendleman, D., Shah, D., et al.: Mobility vla: Multimodal instruc- tion navigation with long-context vlms and topological graphs. arXiv preprint arXiv:2407.07775 (2024) 4

    2024 arXiv

  7. [15]

    Li et al

    Conover, M., Hayes, M., Mathur, A., Xie, J., Wan, J., Shah, S., Ghodsi, A., Wen- dell, P., Zaharia, M., Xin, R.: Free dolly: Introducing the world’s first truly open instructiontuned llm (2023) 10 16 F. Li et al

    2023

  8. [16]

    arXiv preprint arXiv:2506.08566 (2025) 2, 4

    Cui, Y., Xie, L., Zhao, Y., Sun, J., Yin, E.: Generating vision-language naviga- tion instructions incorporated fine-grained alignment annotations. arXiv preprint arXiv:2506.08566 (2025) 2, 4

    2025 arXiv

  9. [17]

    arXiv preprint arXiv:2403.11487 (2024) 4

    Dorbala, V.S., Chowdhury, S., Manocha, D.: Can llms generate human-like wayfinding instructions? towards platform-agnostic embodied instruction synthe- sis. arXiv preprint arXiv:2403.11487 (2024) 4

    2024 arXiv

  10. [18]

    In: European Conference on Computer Vi- sion

    Fan, S., Liu, R., Wang, W., Yang, Y.: Navigation instruction generation with bev perception and large language models. In: European Conference on Computer Vi- sion. pp. 368–387. Springer (2024) 3, 4

    2024

  11. [19]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Fan, S., Liu, R., Wang, W., Yang, Y.: Scene map-based prompt tuning for naviga- tion instruction generation. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 6898–6908 (2025) 4

    2025

  12. [20]

    Communi- cations of the ACM24(6), 381–395 (1981) 7

    Fischler, M.A., Bolles, R.C.: Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Communi- cations of the ACM24(6), 381–395 (1981) 7

    1981

  13. [21]

    Advances in neural information processing sys- tems31(2018) 2, 4

    Fried, D., Hu, R., Cirik, V., Rohrbach, A., Andreas, J., Morency, L.P., Berg- Kirkpatrick, T., Saenko, K., Klein, D., Darrell, T.: Speaker-follower models for vision-and-language navigation. Advances in neural information processing sys- tems31(2018) 2, 4

    2018

  14. [22]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Han, L., Min, H., Hwangbo, G., Choi, J., Seo, P.H.: Dialnav: Multi-turn dialog navigation with a remote guide. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 8514–8523 (2025) 2, 4

    2025

  15. [23]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Han, M., Ma, L., Zhumakhanova, K., Radionova, E., Zhang, J., Chang, X., Liang, X., Laptev, I.: Roomtour3d: Geometry-aware video-instruction tuning for embod- ied navigation. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 27586–27596 (2025) 4, 10

    2025

  16. [24]

    arXiv preprint arXiv:2410.03603 (2024) 4

    Hirose, N., Glossop, C., Sridhar, A., Shah, D., Mees, O., Levine, S.: Lelan: Learning a language-conditioned navigation policy from in-the-wild videos. arXiv preprint arXiv:2410.03603 (2024) 4

    2024 arXiv

  17. [25]

    arXiv preprint arXiv:2501.17403 (2025) 1, 2, 4

    Hong, H., Qiao, Y., Wang, S., Liu, J., Wu, Q.: General scene adaptation for vision- and-language navigation. arXiv preprint arXiv:2501.17403 (2025) 1, 2, 4

    2025 arXiv

  18. [26]

    arXiv preprint arXiv:2506.01946 (2025) 2, 5

    Huang, X., Wu, J., Xie, Q., Han, K.: Mllms need 3d-aware representation super- vision for scene understanding. arXiv preprint arXiv:2506.01946 (2025) 2, 5

    2025

  19. [27]

    In: European Conference on Computer Vision

    Huang, Z., Shangguan, Z., Zhang, J., Bar, G., Boyd, M., Ohn-Bar, E.: Assister: As- sistive navigation via conditional instruction generation. In: European Conference on Computer Vision. pp. 271–289. Springer (2022) 4

    2022

  20. [28]

    arXiv preprint arXiv:1907.05446 (2019) 6

    Ilharco, G., Jain, V., Ku, A., Ie, E., Baldridge, J.: General evaluation for in- struction conditioned navigation using dynamic time warping. arXiv preprint arXiv:1907.05446 (2019) 6

    2019 arXiv

  21. [29]

    In: European Confer- ence on Computer Vision

    Krantz, J., Wijmans, E., Majumdar, A., Batra, D., Lee, S.: Beyond the nav-graph: Vision-and-language navigation in continuous environments. In: European Confer- ence on Computer Vision. pp. 104–120. Springer (2020) 1, 4, 6

    2020

  22. [30]

    arXiv preprint arXiv:2010.07954 (2020) 1, 2, 4, 6, 7

    Ku, A., Anderson, P., Patel, R., Ie, E., Baldridge, J.: Room-across-room: Multilin- gual vision-and-language navigation with dense spatiotemporal grounding. arXiv preprint arXiv:2010.07954 (2020) 1, 2, 4, 6, 7

    2020 arXiv

  23. [31]

    In: Proceedings of the ACM SIGOPS 29th Symposium on Operating Systems Principles (2023) 10 Goal-oriented Navigation Instruction Generation with Tour Video Priors 17

    Kwon, W., Li, Z., Zhuang, S., Sheng, Y., Zheng, L., Yu, C.H., Gonzalez, J.E., Zhang, H., Stoica, I.: Efficient memory management for large language model serv- ing with pagedattention. In: Proceedings of the ACM SIGOPS 29th Symposium on Operating Systems Principles (2023) 10 G...

    2023

  24. [32]

    arXiv preprint arXiv:2502.19024 (2025) 4

    Li, Z., Zhou, G., Hong, H., Shao, Y., Lyu, W., Qiao, Y., Wu, Q.: Ground-level view- point vision-and-language navigation in continuous environments. arXiv preprint arXiv:2502.19024 (2025) 4

    2025 arXiv

  25. [33]

    In: Text sum- marization branches out

    Lin, C.Y.: Rouge: A package for automatic evaluation of summaries. In: Text sum- marization branches out. pp. 74–81 (2004) 8

    2004

  26. [34]

    In: Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision

    Lin, K., Chen, P., Huang, D., Li, T.H., Tan, M., Gan, C.: Learning vision-and- language navigation from youtube videos. In: Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision. pp. 8317–8326 (2023) 4

    2023

  27. [35]

    Liu,F.T.,Ting,K.M.,Zhou,Z.H.:Isolationforest.In:2008eighthieeeinternational conference on data mining. pp. 413–422. IEEE (2008) 7

    2008

  28. [36]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Liu, Y., Zhang, Y., Cai, J., Jiang, X., Hu, Y., Yao, J., Wang, Y., Xie, W.: Lamra: Large multimodal model as your advanced retrieval assistant. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 4015–4025 (2025) 7

    2025

  29. [37]

    arXiv preprint arXiv:2406.04882 (2024) 4

    Long, Y., Cai, W., Wang, H., Zhan, G., Dong, H.: Instructnav: Zero-shot sys- tem for generic instruction navigation in unexplored environment. arXiv preprint arXiv:2406.04882 (2024) 4

    2024 arXiv

  30. [38]

    arXiv preprint arXiv:2402.14830 (2024) 10

    Mitra, A., Khanpour, H., Rosset, C., Awadallah, A.: Orca-math: Unlocking the potential of slms in grade school math. arXiv preprint arXiv:2402.14830 (2024) 10

    2024 arXiv

  31. [39]

    arXiv preprint arXiv:2504.01805 (2025) 10

    Ouyang, K., Liu, Y., Wu, H., Liu, Y., Zhou, H., Zhou, J., Meng, F., Sun, X.: Spacer: Reinforcing mllms in video spatial reasoning. arXiv preprint arXiv:2504.01805 (2025) 10

    2025 arXiv

  32. [40]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Qi, Y., Wu, Q., Anderson, P., Wang, X., Wang, W.Y., Shen, C., Hengel, A.v.d.: Reverie: Remote embodied visual referring expression in real indoor environments. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 9982–9991 (2020) 4

    2020

  33. [41]

    arXiv preprint arXiv:2506.01031 (2025) 8

    Qiao, Y., Hong, H., Lyu, W., An, D., Zhang, S., Xie, Y., Wang, X., Wu, Q.: Navbench: Probing multimodal large language models for embodied navigation. arXiv preprint arXiv:2506.01031 (2025) 8

    2025 arXiv

  34. [42]

    arXiv preprint arXiv:2504.00907 (2025) 4

    Ramrakhya, R., Chang, M., Puig, X., Desai, R., Kira, Z., Mottaghi, R.: Grounding multimodal llms to embodied agents that ask for help with reinforcement learning. arXiv preprint arXiv:2504.00907 (2025) 4

    2025

  35. [43]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Savva, M., Kadian, A., Maksymets, O., Zhao, Y., Wijmans, E., Jain, B., Straub, J., Liu, J., Koltun, V., Malik, J., et al.: Habitat: A platform for embodied ai research. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 9339–9347 (2019) 6

    2019

  36. [44]

    In: Conference on robot learning

    Shah, D., Osiński, B., Levine, S., et al.: Lm-nav: Robotic navigation with large pre- trained models of language, vision, and action. In: Conference on robot learning. pp. 492–504. PMLR (2023) 4

    2023

  37. [45]

    arXiv preprint arXiv:2402.03300 (2024) 9, 4

    Shao, Z., Wang, P., Zhu, Q., Xu, R., Song, J., Bi, X., Zhang, H., Zhang, M., Li, Y., Wu, Y., et al.: Deepseekmath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300 (2024) 9, 4

    2024 arXiv

  38. [46]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Song, X., Chen, W., Liu, Y., Chen, W., Li, G., Lin, L.: Towards long-horizon vision-language navigation: Platform, benchmark and method. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 12078–12088 (2025) 4

    2025

  39. [47]

    arXiv preprint arXiv:1904.04195 (2019) 2, 4

    Tan, H., Yu, L., Bansal, M.: Learning to navigate unseen environments: Back translation with environmental dropout. arXiv preprint arXiv:1904.04195 (2019) 2, 4

    2019 arXiv

  40. [48]

    Li et al

    Team, I.: InternVLA-N1: An open dual-system navigation foundation model with learned latent plans (2025) 9, 13 18 F. Li et al

    2025

  41. [49]

    In: Proceedings of the IEEE/CVF conference on computer vision and pat- tern recognition

    Wang, H., Liang, W., Shen, J., Van Gool, L., Wang, W.: Counterfactual cycle- consistent learning for instruction following and generation in vision-language nav- igation. In: Proceedings of the IEEE/CVF conference on computer vision and pat- tern recognition. pp. 15471–15481 (...

    2022

  42. [50]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Wang, J., Chen, M., Karaev, N., Vedaldi, A., Rupprecht, C., Novotny, D.: Vggt: Visual geometry grounded transformer. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 5294–5306 (2025) 5

    2025

  43. [51]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Wang, S., Montgomery, C., Orbay, J., Birodkar, V., Faust, A., Gur, I., Jaques, N., Waters, A., Baldridge, J., Anderson, P.: Less is more: Generating grounded navi- gation instructions from landmarks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec...

    2022

  44. [52]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Wang, T., Cheng, C., Wang, L., Chen, S., Zhao, W.: Himtok: Learning hierarchical mask tokens for image segmentation with large multimodal model. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 23267–23278 (2025) 5, 10

    2025

  45. [53]

    Advances in Neural Information Processing Systems37, 121475–121499 (2024) 10

    Wang, W., Lv, Q., Yu, W., Hong, W., Qi, J., Wang, Y., Ji, J., Yang, Z., Zhao, L., XiXuan, S., et al.: Cogvlm: Visual expert for pretrained language models. Advances in Neural Information Processing Systems37, 121475–121499 (2024) 10

    2024

  46. [54]

    arXiv preprint arXiv:2411.10442 (2024) 10, 5

    Wang, W., Chen, Z., Wang, W., Cao, Y., Liu, Y., Gao, Z., Zhu, J., Zhu, X., Lu, L., Qiao, Y., et al.: Enhancing the reasoning ability of multimodal large language models via mixed preference optimization. arXiv preprint arXiv:2411.10442 (2024) 10, 5

    2024 arXiv

  47. [55]

    5: Advancing open-source multimodal models in versatility, reasoning, and efficiency

    Wang, W., Gao, Z., Gu, L., Pu, H., Cui, L., Wei, X., Liu, Z., Jing, L., Ye, S., Shao, J., et al.: Internvl3. 5: Advancing open-source multimodal models in versatility, reasoning, and efficiency. arXiv preprint arXiv:2508.18265 (2025) 5, 10, 11, 6, 7

    2025 arXiv

  48. [56]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Wang, X., Wang, W., Shao, J., Yang, Y.: Lana: A language-capable navigator for instruction following and generation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 19048–19058 (2023) 2, 4

    2023

  49. [57]

    arXiv preprint arXiv:2502.11142 (2025) 4

    Wang, Z., Zhu, Y., Lee, G.H., Fan, Y.: Navrag: Generating user demand instruc- tions for embodied navigation through retrieval-augmented llm. arXiv preprint arXiv:2502.11142 (2025) 4

    2025 arXiv

  50. [58]

    arXiv preprint arXiv:2412.08467 (2024) 4

    Wang, Z., Li, J., Hong, Y., Li, S., Li, K., Yu, S., Wang, Y., Qiao, Y., Wang, Y., Bansal, M., et al.: Bootstrapping language-guided navigation learning with self- refining data flywheel. arXiv preprint arXiv:2412.08467 (2024) 4

    2024 arXiv

  51. [59]

    In: Proceed- ings of the IEEE/CVF international conference on computer vision

    Wang, Z., Li, J., Hong, Y., Wang, Y., Wu, Q., Bansal, M., Gould, S., Tan, H., Qiao, Y.: Scaling data generation in vision-and-language navigation. In: Proceed- ings of the IEEE/CVF international conference on computer vision. pp. 12009– 12020 (2023) 2, 4

    2023

  52. [60]

    arXiv preprint arXiv:2512.08186 (2025) 9, 13

    Wei, M., Wan, C., Peng, J., Yu, X., Yang, Y., Feng, D., Cai, W., Zhu, C., Wang, T., Pang, J., et al.: Ground slow, move fast: A dual-system foundation model for generalizable vision-and-language navigation. arXiv preprint arXiv:2512.08186 (2025) 9, 13

    2025

  53. [61]

    arXiv preprint arXiv:2507.05240 (2025) 4

    Wei,M.,Wan,C.,Yu,X.,Wang,T.,Yang,Y.,Mao,X.,Zhu,C.,Cai,W.,Wang,H., Chen, Y., et al.: Streamvln: Streaming vision-and-language navigation via slowfast context modeling. arXiv preprint arXiv:2507.05240 (2025) 4

    2025 arXiv

  54. [62]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Wijmans, E., Datta, S., Maksymets, O., Das, A., Gkioxari, G., Lee, S., Essa, I., Parikh, D., Batra, D.: Embodied question answering in photorealistic environments with point cloud perception. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition...

    2019

  55. [63]

    arXiv preprint arXiv:2508.09547 (2025) 2, 3, 5

    Wu, F., Dong, Y., Cheng, Z.Q., Dai, Y., Chen, G., Wang, H., Dai, Q., Haupt- mann, A.G.: Govig: Goal-conditioned visual navigation instruction generation. arXiv preprint arXiv:2508.09547 (2025) 2, 3, 5

    2025 arXiv

  56. [64]

    In: Pro- ceedings of the Computer Vision and Pattern Recognition Conference

    Yang, J., Yang, S., Gupta, A.W., Han, R., Fei-Fei, L., Xie, S.: Thinking in space: How multimodal large language models see, remember, and recall spaces. In: Pro- ceedings of the Computer Vision and Pattern Recognition Conference. pp. 10632– 10643 (2025) 2, 5, 12

    2025

  57. [65]

    In: 2024 IEEE International Con- ference on Robotics and Automation (ICRA)

    Yokoyama, N., Ha, S., Batra, D., Wang, J., Bucher, B.: Vlfm: Vision-language frontier maps for zero-shot semantic navigation. In: 2024 IEEE International Con- ference on Robotics and Automation (ICRA). pp. 42–48. IEEE (2024) 2

    2024

  58. [66]

    In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    Yokoyama, N., Ramrakhya, R., Das, A., Batra, D., Ha, S.: Hm3d-ovon: A dataset and benchmark for open-vocabulary object goal navigation. In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). pp. 5543–

    2024

  59. [67]

    arXiv preprint arXiv:2309.12284 (2023) 10

    Yu, L., Jiang, W., Shi, H., Yu, J., Liu, Z., Zhang, Y., Kwok, J.T., Li, Z., Weller, A., Liu, W.: Metamath: Bootstrap your own mathematical questions for large language models. arXiv preprint arXiv:2309.12284 (2023) 10

    2023 arXiv

  60. [68]

    arXiv preprint arXiv:2309.05653 (2023) 10

    Yue, X., Qu, X., Zhang, G., Fu, Y., Huang, W., Sun, H., Su, Y., Chen, W.: Mam- moth: Building math generalist models through hybrid instruction tuning. arXiv preprint arXiv:2309.05653 (2023) 10

    2023 arXiv

  61. [69]

    arXiv preprint arXiv:2412.06224 (2024) 4

    Zhang, J., Wang, K., Wang, S., Li, M., Liu, H., Wei, S., Wang, Z., Zhang, Z., Wang, H.: Uni-navid: A video-based vision-language-action model for unifying embodied navigation tasks. arXiv preprint arXiv:2412.06224 (2024) 4

    2024 arXiv

  62. [70]

    arXiv preprint arXiv:2402.15852 (2024) 4

    Zhang, J., Wang, K., Xu, R., Zhou, G., Hong, Y., Fang, X., Wu, Q., Zhang, Z., Wang, H.: Navid: Video-based vlm plans the next step for vision-and-language navigation. arXiv preprint arXiv:2402.15852 (2024) 4

    2024 arXiv

  63. [71]

    arXiv preprint arXiv:2502.13451 (2025) 4

    Zhang, L., Hao, X., Xu, Q., Zhang, Q., Zhang, X., Wang, P., Zhang, J., Wang, Z., Zhang, S., Xu, R.M.: A novel memory representation via annotated semantic maps for vlm-based vision-and-language navigation. arXiv preprint arXiv:2502.13451 (2025) 4

    2025 arXiv

  64. [72]

    arXiv preprint arXiv:2502.14254 (2025) 4

    Zhang, L., Liu, Y., Zhang, Z., Aghaei, M., Hu, Y., Gu, H., Alomrani, M.A., Bravo, D.G.A., Karimi, R., Hamidizadeh, A., et al.: Mem2ego: Empowering vision- language models with global-to-ego memory for long-horizon embodied navigation. arXiv preprint arXiv:2502.14254 (2025) 4

    2025 arXiv

  65. [73]

    arXiv preprint arXiv:2407.07035 (2024) 2

    Zhang, Y., Ma, Z., Li, J., Qiao, Y., Wang, Z., Chai, J., Wu, Q., Bansal, M., Ko- rdjamshidi, P.: Vision-and-language navigation today and tomorrow: A survey in the era of foundation models. arXiv preprint arXiv:2407.07035 (2024) 2

    2024 arXiv

  66. [74]

    arXiv preprint arXiv:2101.10504 (2021) 8

    Zhao, M., Anderson, P., Jain, V., Wang, S., Ku, A., Baldridge, J., Ie, E.: On the evaluation of vision-and-language navigation instructions. arXiv preprint arXiv:2101.10504 (2021) 8

    2021 arXiv

  67. [75]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Zheng, D., Huang, S., Zhao, L., Zhong, Y., Wang, L.: Towards learning a generalist model for embodied navigation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 13624–13634 (2024) 1, 2, 4

    2024

  68. [76]

    arXiv preprint arXiv:2402.14658 (2024) 10

    Zheng, T., Zhang, G., Shen, T., Liu, X., Lin, B.Y., Fu, J., Chen, W., Yue, X.: Opencodeinterpreter: Integrating code generation with execution and refinement. arXiv preprint arXiv:2402.14658 (2024) 10

    2024 arXiv

  69. [77]

    In: European Confer- ence on Computer Vision

    Zhou, G., Hong, Y., Wang, Z., Wang, X.E., Wu, Q.: Navgpt-2: Unleashing naviga- tional reasoning capability for large vision-language models. In: European Confer- ence on Computer Vision. pp. 260–278. Springer (2024) 2, 4, 5 20 F. Li et al

    2024

  70. [78]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Zhou, S., Vilesov, A., He, X., Wan, Z., Zhang, S., Nagachandra, A., Chang, D., Chen, D., Wang, X.E., Kadambi, A.: Vlm4d: Towards spatiotemporal awareness in vision language models. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 8600–8612 (2025) 2, 5

    2025

  71. [79]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Zhu, F., Liang, X., Zhu, Y., Yu, Q., Chang, X., Liang, X.: Soon: Scenario oriented object navigation with graph-based exploration. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 12689–12699 (2021) 2, 4

    2021

  72. [80]

    Move forward 100 cm

    Zhu, J., Wang, W., Chen, Z., Liu, Z., Ye, S., Gu, L., Tian, H., Duan, Y., Su, W., Shao, J., et al.: Internvl3: Exploring advanced training and test-time recipes for open-source multimodal models. arXiv preprint arXiv:2504.10479 (2025) 10, 6, 7 Goal-oriented Navigation Instruct...

    2025 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.