Pith. sign in

REVIEW 3 major objections 3 minor 38 references

$\mathcal{P}^3$: Toward Versatile Embodied Agents

T0 review · 3 major / 3 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read P^3 claims that embodied agents can perceive the environment directly, skip tool feedback, and dynamically reprioritize multi-task plans in real time.

desk verdict The P^3 abstract promises an embodied-agent framework, but the submitted full text is an unrelated 3DGS benchmark; the actual paper is missing. read the letter →

arxiv 2508.07033 v2 pith:UVFSMHOD submitted 2025-08-09 cs.RO

Shengli Zhou , Xiangchen Wang , Jinrui Zhang , Ruozai Tian , Rongtao Xu , Guanhua Chen , Feng Zheng This is my paper
classification cs.RO
keywords embodiedagentsreal-timeperceptiondynamicschedulingtoolusemulti-taskplanningtaskdependenciesfeedback-freeoperationdeployment
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 argues that embodied agents do not need to wait for tools to report back. It introduces $P^{3}$, a unified framework that couples real-time perception with dynamic scheduling, so an agent can actively inspect the environment, plug in tools without feedback loops, and reorder its to-do list by urgency and dependency. The intended payoff is a general-purpose embodied agent that transfers from benchmarks to live deployment. The abstract promises extensive real-world experiments, but the full text provided with this submission is a different paper on 3DGS video quality evaluation, so those experiments are not present to inspect.

What carries the argument

$P^{3}$ itself is the central object: a unified framework joining a real-time perception module with a dynamic scheduling module. The perception module actively fetches task-relevant state from the environment instead of depending on tool-returned feedback; the scheduling module prioritizes urgent tasks and rewrites task order as dependencies shift. The load-bearing move is that environment perception can stand in for tool feedback.

What would settle it

Run the abstract's real-world multi-task experiments and compare an agent with tool feedback disabled against one with feedback enabled over long episodes: if task success degrades, state estimates drift, or tool misuse goes undetected, the central claim fails. The full text provided is a different paper (3DGS-VBench), so the claimed experiments are not currently available for this check.

Watch

Extended reading notes

Core claim

The central claim is that three bottlenecks—dynamic environment perception, open-ended tool use, and multi-task planning—can be handled in one system by replacing tool-feedback loops with active perception and by making the scheduler dependency- and urgency-aware. Concretely, $P^{3}$ is claimed to perceive task-relevant information directly from the environment, use tools without requiring feedback from the tool, and plan multi-task execution that prioritizes urgent tasks and dynamically adjusts order when dependencies change. If this works, the reliance on tool feedback that previous approaches need is removed, and embodied agents become more transferable across environments and tasks.

Load-bearing premise

For the central claim to hold, actively sensing the environment must supply enough reliable information about task and tool state that omitting tool feedback does not let errors go unnoticed or accumulate over time.

Editorial extensions

If this is right

  • An agent can adopt a new tool without any tool-specific feedback protocol, because it watches the effect of the tool rather than trusting the tool's own report.
  • Multi-task execution can be interrupted by urgent tasks and reordered around dependencies, which is a step beyond fixed task lists and static scheduling.
  • If the framework is correct, error accumulation from chained tool feedback is reduced, since state is re-read from the environment continuously.
  • The same perception-plus-scheduling core could transfer across different physical environments, which is the paper's stated route to practical deployment.

Reading between the lines

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

  • If feedback-free perception is the load-bearing component, the characteristic failure mode shifts from wrong tool output to unnoticed tool misuse: a tool could act correctly but leave no visible trace, and the agent would be blind to it. A testable extension would measure success on tasks where tool effects are imperceptible in the agent's sensor stream.
  • The dependency-aware scheduler is naturally extensible to multi-agent settings: an urgent task for one agent can be treated as a newly injected dependency for another, so P^3-style scheduling could become a coordination primitive rather than only a planning module.
  • The abstract's emphasis on real-time perception implies a cost trade-off that the paper does not state explicitly: perception frequency and state-estimation accuracy now carry the load formerly borne by tool feedback, so the framework's robustness should be probed by varying sensor update rates.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 3 minor

Summary. The manuscript metadata (title, abstract, and GitHub link) describes P^3, a proposed unified framework for embodied agents that integrates real-time perception and dynamic scheduling. The abstract claims that P^3 enables agents to actively perceive task-relevant information, use tools without feedback requirements, and dynamically reorder multi-task execution, with 'extensive real-world experiments' demonstrating transferable, general-purpose embodied agents. However, the supplied full text is a completely different paper: arXiv:2508.07038v1, titled '3DGS-VBench: A Comprehensive Video Quality Evaluation Benchmark for 3DGS Compression.' This full text contains no mention of P^3, no embodied-agent architecture, no equations, no experiments on tool use or scheduling, and no evaluation. The submitted artifact therefore does not contain any of the evidence needed to support the abstract's claims.

Significance. If the P^3 framework had been properly described and evaluated, its central claims would be significant for embodied AI: replacing tool-feedback loops with active environment perception, and adding dynamic scheduling for multi-task execution, could address real-world deployment gaps. However, the manuscript as submitted is unverifiable. The abstract is not even accompanied by a matching full text, so no architecture, no derivations, no baselines, no ablation studies, and no quantitative results are available for scrutiny. This is not a disagreement with consensus; it is a complete absence of the object of review.

major comments (3)
  1. [Full text (title, Abstract, Sections I–III)] The full text is arXiv:2508.07038v1, '3DGS-VBench: A Comprehensive Video Quality Evaluation Benchmark for 3DGS Compression.' It is a VQA benchmark paper for 3D Gaussian Splatting. It contains zero content about P^3, embodied agents, real-time perception, tool feedback, multi-task scheduling, or the claimed experiments. The abstract's central assertions—especially 'extensive real-world experiments' and 'highly transferable, general-purpose embodied agents'—are therefore entirely unsupported by any visible protocol, metric, baseline, or number.
  2. [Abstract, 'without feedback requirements'] The load-bearing design claim that the agent can 'plug and utilize tools without feedback requirements' is a strong negative claim. Substantiation would require long-horizon trajectory studies, failure-rate measurements, state-estimation drift analysis, and ablations against tool-feedback baselines. None of this is present. The submitted full text provides no mechanism or experiment to rule out error accumulation, tool misuse, or missed state changes when feedback loops are removed.
  3. [Abstract vs. full text (internal consistency)] The manuscript is internally inconsistent: the abstract and associated metadata describe P^3, while the full text describes a different paper with a different author list, different contributions, and different references. This mismatch blocks any correctness assessment of the claimed framework. A referee cannot evaluate a paper whose body is a different paper.
minor comments (3)
  1. [Title/abstract metadata] The title, author list, abstract, and GitHub link reference P^3, but the full text is from a different paper with a different title and author list. This is not a minor typo; it is a fundamental metadata mismatch that should be resolved before any further review.
  2. [Abstract, 'Code and data are available'] The GitHub link https://github.com/fz-zsl/P3 may contain the P^3 project, but the submitted full text gives no description of that repository's contents, and no code or data is described in the paper body. The link alone cannot substitute for the missing experiments.
  3. [References] The references in the full text are all for 3DGS and VQA (e.g., [1] Kerbl et al., [2] Mildenhall et al., [12]–[18] QA datasets). None of them relate to embodied agents, perception, tool use, or scheduling. If the intended P^3 paper were attached, a relevant related-work section and reference list would be needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the P^3 abstract contains no derivation chain, and the supplied full text is an unrelated benchmark paper; no load-bearing step reduces to its own inputs.

full rationale

The artifact under review is the abstract of arXiv:2508.07033 (P^3), but the supplied full text is arXiv:2508.07038v1 (3DGS-VBench), a different paper on video quality assessment for 3D Gaussian Splatting compression. The P^3 abstract contains no equations, fitted parameters, or derivation chain. Its central claims—active environment perception, tool use without feedback, dynamic multi-task scheduling—are stated programmatically, not derived from any formal framework in the provided text. There is therefore no equation or construction by which a prediction reduces to its own inputs, and no fitted quantity is renamed as a prediction. The 3DGS-VBench full text is a dataset and benchmark construction paper: MOS scores come from subjective human annotations, and the benchmark evaluates existing compression algorithms and QA metrics. Its self-citations to prior work by overlapping authors (e.g., references [15] and [16]) appear only in related-work context as descriptions of existing datasets, not as load-bearing evidence for the paper's own benchmark conclusions. No uniqueness theorem, ansatz, or known result is imported via self-citation to force the paper's choices. The mismatch between the P^3 abstract and the supplied full text is a serious correctness and verifiability concern, but it is not a circularity pattern: the central claims are unsupported rather than self-referential. Under the hard rule requiring a specific reduction to be quoted, no circular step can be identified, so the score is 0.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

No equations, fitted constants, or hand-chosen values appear in the abstract, so the free-parameter ledger is empty at this level of review. The two domain assumptions above are load-bearing for the abstract's claims. A complete ledger requires the body text, which was not supplied: the submitted full text is the 3DGS-VBench paper, not the P^3 manuscript.

assumptions (2)
  • domain assumption Active perception without tool feedback is sufficient for accurate task-relevant state estimation in dynamic environments.
    The abstract's design claim that agents can 'utilize tools without feedback requirements' depends on this premise; if it fails, long-horizon error accumulation would return and the framework's stated advantage disappears.
  • domain assumption Tool-agent feedback reliance is the binding constraint limiting prior embodied agents.
    The abstract motivates P^3 entirely by this limitation, but no comparative evidence appears in the abstract, so the premise is asserted, not demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of $\mathcal{P}^3$: Toward Versatile Embodied Agents." pith.science (2026). https://pith.science/paper/UVFSMHOD

@misc{pith2026250807033,
  author       = {Pith},
  title        = {Pith review of: $\mathcalP^3$: Toward Versatile Embodied Agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UVFSMHOD}},
  note         = {Machine review of arXiv:2508.07033}
}
abstract

Embodied agents have shown promising generalization capabilities across diverse physical environments, making them essential for a wide range of real-world applications. However, building versatile embodied agents poses critical challenges due to three key issues: dynamic environment perception, open-ended tool usage, and complex multi-task planning. Most previous works rely solely on feedback from tool agents to perceive environmental changes and task status, which limits adaptability to real-time dynamics, causes error accumulation, and restricts tool flexibility. Furthermore, multi-task scheduling has received limited attention, primarily due to the inherent complexity of managing task dependencies and balancing competing priorities in dynamic and complex environments. To overcome these challenges, we introduce $\mathcal P^3$, a unified framework that integrates real-time perception and dynamic scheduling. Specifically, $\mathcal P^3$ enables agents to perceive task-relevant information actively from the environment, plug and utilize tools without feedback requirements, and plan multi-task execution by prioritizing urgent tasks and dynamically adjusting task order based on dependencies. Extensive real-world experiments show that our approach bridges the gap between benchmarks and practical deployment, delivering highly transferable, general-purpose embodied agents. Code and data are available at https://github.com/fz-zsl/P3.

Figures

Figures reproduced from arXiv: 2508.07033 by the authors.

Figure 1
Figure 1. 11 selected scenes in 3DGS-VBench: scenes 1-5 depict outdoor scenes, while scenes 6-11 depict indoor scenes. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. (a), where the labels match the scene numbers in [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. MOS vs. Storage (MB) for LightGaussian with diverse [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

38 extracted references · 30 canonical work pages

  1. [1]

    3d gaussian splatting for real-time radiance field rendering,

    B. Kerbl, G. Kopanas, T. Leimk ¨uhler, and G. Drettakis, “3d gaussian splatting for real-time radiance field rendering,” ACM Transactions on Graphics, vol. 42, no. 4, 2023

    work page 2023

  2. [2]

    Nerf: representing scenes as neural radiance fields for view synthesis,

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, “Nerf: representing scenes as neural radiance fields for view synthesis,” Commun. ACM, vol. 65, no. 1, p. 99–106, 2021

    work page 2021

  3. [3]

    Scaffold-gs: Structured 3d gaussians for view-adaptive rendering,

    T. Lu, M. Yu, L. Xu, Y . Xiangli, L. Wang, D. Lin, and B. Dai, “Scaffold-gs: Structured 3d gaussians for view-adaptive rendering,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 20654–20664, 2024

    work page 2024

  4. [4]

    Hac: Hash-grid assisted context for 3d gaussian splatting compression,

    Y . Chen, Q. Wu, W. Lin, M. Harandi, and J. Cai, “Hac: Hash-grid assisted context for 3d gaussian splatting compression,” in European Conference on Computer Vision (ECCV) , pp. 422–438, Springer, 2024

    work page 2024

  5. [5]

    Light- gaussian: Unbounded 3d gaussian compression with 15x reduction and 200+ fps,

    Z. Fan, K. Wang, K. Wen, Z. Zhu, D. Xu, Z. Wang, et al. , “Light- gaussian: Unbounded 3d gaussian compression with 15x reduction and 200+ fps,” Advances in neural information processing systems , vol. 37, pp. 140138–140158, 2024

    work page 2024

  6. [6]

    Compgs: Smaller and faster gaussian splatting with vector quantization,

    K. Navaneet, K. Pourahmadi Meibodi, S. Abbasi Koohpayegani, and H. Pirsiavash, “Compgs: Smaller and faster gaussian splatting with vector quantization,” in European Conference on Computer Vision , pp. 330–349, Springer, 2024

    work page 2024

  7. [7]

    Compressed 3d gaussian splatting for accelerated novel view synthesis,

    S. Niedermayr, J. Stumpfegger, and R. Westermann, “Compressed 3d gaussian splatting for accelerated novel view synthesis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion (CVPR), pp. 10349–10358, June 2024

    work page 2024

  8. [8]

    Compact 3d gaussian splatting for static and dynamic radiance fields,

    J. C. Lee, D. Rho, X. Sun, J. H. Ko, and E. Park, “Compact 3d gaussian splatting for static and dynamic radiance fields,” arXiv preprint arXiv:2408.03822, 2024

    arXiv 2024

Show all 38 references
  1. [9]

    Eagles: Efficient accelerated 3d gaussians with lightweight encodings,

    S. Girish, K. Gupta, and A. Shrivastava, “Eagles: Efficient accelerated 3d gaussians with lightweight encodings,” in European Conference on Computer Vision, pp. 54–71, Springer, 2024

    2024

  2. [10]

    Compgs: Efficient 3d scene representation via compressed gaussian splatting,

    X. Liu, X. Wu, P. Zhang, S. Wang, Z. Li, and S. Kwong, “Compgs: Efficient 3d scene representation via compressed gaussian splatting,” in Proceedings of the 32nd ACM International Conference on Multimedia , pp. 2936–2944, 2024

    2024

  3. [11]

    Hybridgs: High-efficiency gaussian splatting data compression using dual- channel sparse representation and point cloud encoder,

    Q. Yang, L. Yang, G. Van Der Auwera, and Z. Li, “Hybridgs: High-efficiency gaussian splatting data compression using dual- channel sparse representation and point cloud encoder,” arXiv preprint arXiv:2505.01938, 2025

    2025 arXiv

  4. [12]

    Nerf-qa: Neural radiance fields quality assessment database,

    P. Martin, A. Rodrigues, J. Ascenso, and M. P. Queluz, “Nerf-qa: Neural radiance fields quality assessment database,” in 2023 15th International Conference on Quality of Multimedia Experience (QoMEX) , pp. 107– 110, 2023

    2023

  5. [13]

    Nerf view synthesis: Subjective quality assessment and objective metrics evaluation,

    P. Martin, A. Rodrigues, J. Ascenso, and M. Paula Queluz, “Nerf view synthesis: Subjective quality assessment and objective metrics evaluation,” IEEE Access, vol. 13, pp. 26–41, 2025

    2025

  6. [14]

    Perceptual quality assessment of nerf and neural view synthesis methods for front-facing views,

    H. Liang, T. Wu, P. Hanji, F. Banterle, H. Gao, R. Mantiuk, and C. ¨Oztireli, “Perceptual quality assessment of nerf and neural view synthesis methods for front-facing views,” inComputer Graphics Forum, vol. 43, p. e15036, Wiley Online Library, 2024

    2024

  7. [15]

    Explicit-nerf-qa: A quality assessment database for explicit nerf model compression,

    Y . Xing, Q. Yang, K. Yang, Y . Xu, and Z. Li, “Explicit-nerf-qa: A quality assessment database for explicit nerf model compression,” in 2024 IEEE International Conference on Visual Communications and Image Processing (VCIP) , pp. 1–5, 2024

    2024

  8. [16]

    A benchmark for gaussian splatting compression and quality assessment study,

    Q. Yang, K. Yang, Y . Xing, Y . Xu, and Z. Li, “A benchmark for gaussian splatting compression and quality assessment study,” in Proceedings of the 6th ACM International Conference on Multimedia in Asia , pp. 1–8, 2024

    2024

  9. [17]

    Gs-qa: Com- prehensive quality assessment benchmark for gaussian splatting view synthesis,

    P. Martin, A. Rodrigues, J. Ascenso, and M. P. Queluz, “Gs-qa: Com- prehensive quality assessment benchmark for gaussian splatting view synthesis,” arXiv preprint arXiv:2502.13196 , 2025

    2025 arXiv

  10. [18]

    Evaluating human perception of novel view synthesis: Subjective quality assessment of gaussian splatting and nerf in dynamic scenes,

    Y . Zhang, J. Maraval, Z. Zhang, N. Ramin, S. Tian, and L. Zhang, “Evaluating human perception of novel view synthesis: Subjective quality assessment of gaussian splatting and nerf in dynamic scenes,” arXiv preprint arXiv:2501.08072 , 2025

    2025 arXiv

  11. [19]

    Mip-nerf 360: Unbounded anti-aliased neural radiance fields,

    J. T. Barron, B. Mildenhall, D. Verbin, P. P. Srinivasan, and P. Hedman, “Mip-nerf 360: Unbounded anti-aliased neural radiance fields,” in Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR), pp. 5470–5479, 2022

    2022

  12. [20]

    Tanks and temples: Benchmarking large-scale scene reconstruction,

    A. Knapitsch, J. Park, Q.-Y . Zhou, and V . Koltun, “Tanks and temples: Benchmarking large-scale scene reconstruction,” ACM Transactions on Graphics (ToG), vol. 36, no. 4, pp. 1–13, 2017

    2017

  13. [21]

    Deep blending for free-viewpoint image-based rendering,

    P. Hedman, J. Philip, T. Price, J.-M. Frahm, G. Drettakis, and G. Bros- tow, “Deep blending for free-viewpoint image-based rendering,” ACM Transactions on Graphics (ToG), vol. 37, no. 6, pp. 1–15, 2018

    2018

  14. [22]

    Pku-dymvhumans: A multi-view video benchmark for high- fidelity dynamic human modeling,

    X. Zheng, L. Liao, X. Li, J. Jiao, R. Wang, F. Gao, S. Wang, and R. Wang, “Pku-dymvhumans: A multi-view video benchmark for high- fidelity dynamic human modeling,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pp. 22530– 22540, 2024

    2024

  15. [23]

    Subjective video quality assessment methods for multimedia applications,

    P. ITU-T RECOMMENDATION, “Subjective video quality assessment methods for multimedia applications,” 1999

    1999

  16. [24]

    Image quality assess- ment: From error visibility to structural similarity,

    Z. Wang, A. Bovik, H. Sheikh, and E. Simoncelli, “Image quality assess- ment: From error visibility to structural similarity,” IEEE Transactions on Image Processing , vol. 13, no. 4, pp. 600–612, 2004

    2004

  17. [25]

    The unreasonable effectiveness of deep features as a perceptual metric,

    R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang, “The unreasonable effectiveness of deep features as a perceptual metric,” in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, pp. 586–595, 2018

    2018

  18. [26]

    Image quality assess- ment: Unifying structure and texture similarity,

    K. Ding, K. Ma, S. Wang, and E. P. Simoncelli, “Image quality assess- ment: Unifying structure and texture similarity,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 44, no. 5, pp. 2567– 2581, 2022

    2022

  19. [27]

    Image information and visual quality,

    H. Sheikh and A. Bovik, “Image information and visual quality,” IEEE Transactions on Image Processing , vol. 15, no. 2, pp. 430–444, 2006

    2006

  20. [28]

    Fsim: A feature similarity index for image quality assessment,

    L. Zhang, L. Zhang, X. Mou, and D. Zhang, “Fsim: A feature similarity index for image quality assessment,” IEEE Transactions on Image Processing, vol. 20, no. 8, pp. 2378–2386, 2011

    2011

  21. [29]

    Information content weighting for perceptual image quality assessment,

    Z. Wang and Q. Li, “Information content weighting for perceptual image quality assessment,” IEEE Transactions on Image Processing , vol. 20, no. 5, pp. 1185–1198, 2011

    2011

  22. [30]

    Multiscale structural similarity for image quality assessment,

    Z. Wang, E. Simoncelli, and A. Bovik, “Multiscale structural similarity for image quality assessment,” in The Thrity-Seventh Asilomar Confer- ence on Signals, Systems & Computers, 2003 , vol. 2, pp. 1398–1402 V ol.2, 2003

    2003

  23. [31]

    Exploring clip for assessing the look and feel of images,

    J. Wang, K. C. Chan, and C. C. Loy, “Exploring clip for assessing the look and feel of images,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, pp. 2555–2563, 2023

    2023

  24. [32]

    No-reference image quality assessment in the spatial domain,

    A. Mittal, A. K. Moorthy, and A. C. Bovik, “No-reference image quality assessment in the spatial domain,” IEEE Transactions on Image Processing, vol. 21, no. 12, pp. 4695–4708, 2012

    2012

  25. [33]

    Exploring video quality assessment on user generated contents from aesthetic and technical perspectives,

    H. Wu, E. Zhang, L. Liao, C. Chen, J. Hou, A. Wang, W. Sun, Q. Yan, and W. Lin, “Exploring video quality assessment on user generated contents from aesthetic and technical perspectives,” inProceedings of the IEEE/CVF International Conference on Computer Vision , pp. 20144– 20154, 2023

    2023

  26. [34]

    Fast-vqa: Efficient end-to-end video quality assessment with fragment sampling,

    H. Wu, C. Chen, J. Hou, L. Liao, A. Wang, W. Sun, Q. Yan, and W. Lin, “Fast-vqa: Efficient end-to-end video quality assessment with fragment sampling,” in European conference on computer vision , pp. 538–554, Springer, 2022

    2022

  27. [35]

    A deep learning based no- reference quality assessment model for ugc videos,

    W. Sun, X. Min, W. Lu, and G. Zhai, “A deep learning based no- reference quality assessment model for ugc videos,” in Proceedings of the 30th ACM International Conference on Multimedia , pp. 856–865, 2022

    2022

  28. [36]

    Quality assessment of in-the-wild videos,

    D. Li, T. Jiang, and M. Jiang, “Quality assessment of in-the-wild videos,” in Proceedings of the 27th ACM international conference on multimedia, pp. 2351–2359, 2019

    2019

  29. [37]

    Q-align: Teaching lmms for visual scoring via discrete text-defined levels,

    H. Wu, Z. Zhang, W. Zhang, C. Chen, L. Liao, C. Li, Y . Gao, A. Wang, E. Zhang, W. Sun, et al. , “Q-align: Teaching lmms for visual scoring via discrete text-defined levels,” arXiv preprint arXiv:2312.17090, 2023

    2023 arXiv

  30. [38]

    Mped: Quantifying point cloud distortion based on multiscale potential energy discrepancy,

    Q. Yang, Y . Zhang, S. Chen, Y . Xu, J. Sun, and Z. Ma, “Mped: Quantifying point cloud distortion based on multiscale potential energy discrepancy,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 5, pp. 6037–6054, 2023

    2023

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.