Pith. sign in

REVIEW 4 major objections 7 minor 36 references

This paper claims that 3D root skeletons can be reconstructed from as few as three multi-angle images by detecting, matching, and triangulating lateral roots, and that the recovered architecture is close to ground truth.

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 ↗

A multi-view computer vision pipeline detects and matches lateral roots in images, triangulates them, and refines the result with bundle adjustment to reconstruct 3D root skeletons from a few views, evaluated on a custom sweet potato dataset.

T0 review reviewed 2026-08-05 challenge →

load-bearing objection Promising pipeline for 3D root skeletons, but the evaluation is on clean renders, not real photos, so the headline claim overreaches. the 4 major comments →

arxiv 2508.08094 v1 pith:OODAGART submitted 2025-08-11 cs.CV

3D Plant Root Skeleton Detection and Extraction

Jiakai Lin , Jinchang Zhang , Ge Jin , Wenzhan Song , Tianming Liu , Guoyu Lu This is my paper
classification cs.CV
keywords 3D root skeletonplant phenotypinglateral root detectionmulti-view triangulationbundle adjustmentsweet potato rootsbreeding robots
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

This paper aims to establish that dense, texture-poor plant root systems can be converted into 3D skeletons from as few as three to ten ordinary images, without depth sensors or synchronized camera rigs. The key move is to detect, match, and triangulate whole lateral roots instead of reconstructing point clouds: a convolutional detector finds lateral roots and their endpoints, keypoint votes link the same root across views, and weighted triangulation turns matched roots into 3D lines. A differentiable bundle-adjustment layer then refines skeleton points and camera poses, supervised by the fact that angles between roots are invariant in 3D. The main root is assembled by connecting lateral-root starting points in growth order, which also separates the two root types. The pipeline is validated on a custom set of 400 scanned sweet-potato root meshes, with 40 held out for testing, reporting higher 3D lateral-root precision and recall than depth-estimation-plus-skeletonization baselines.

Core claim

The paper's central claim is that 3D root skeletons can be reconstructed from as few as three to ten images by detecting, matching, and triangulating lateral roots instead of dense point-cloud reconstruction. Lateral roots are located as objects with start and end keypoints; matching across views is decided by keypoint votes between bounding boxes; two-way triangulation is weighted by reprojection error; and a learnable bundle-adjustment layer, supervised by the invariance of root angles, refines skeleton and camera poses. Main-root topology is obtained by connecting lateral-root origins in growth order. On 40 held-out sweet-potato root systems, the reported 3D lateral-root precision is 0.77

What carries the argument

The load-bearing object is the lateral root treated as a detected and matched unit rather than a set of independent points. A convolutional detector outputs boxes plus start/end keypoints for each lateral root; a keypoint-voting step builds a match-score matrix to associate boxes across views; weighted triangulation fuses two-view reconstructions; and a differentiable Skeleton Bundle Adjustment layer minimizes reprojection error and angle discrepancies, predicting the Levenberg-Marquardt damping factor with attention so the whole optimization is trainable by backpropagation. The main root is assembled by a row-wise growth-simulation matrix that connects lateral-root origins in order.

Load-bearing premise

The load-bearing premise is that projection images rendered from manually denoised 3D meshes, with camera poses known exactly, stand in for real photographs of roots growing in soil.

What would settle it

Photograph real root systems with a handheld camera and unknown poses, or render the same models with soil occlusion and perturbed cameras, then compare recovered skeletons against CT-derived ground truth. If lateral-root precision and recall fall to depth-estimation baseline levels, the clean-mesh known-pose rendering is what produces the reported accuracy.

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

If this is right

  • Root architecture phenotyping — root counts, branching order, angles, lengths — becomes possible from a few handheld or robot-arm images instead of multi-camera or destructive measurements.
  • Primary and lateral roots are separated automatically, so trait measurements can be reported per root type.
  • The angle-consistency self-supervision removes the need for dense 3D ground truth during skeleton refinement.
  • With three views enough for reconstruction, capture can be done by a moving camera, simplifying breeding-robot integration.
  • Breeding robots could use the resulting skeleton as direct input for selecting plants with favorable root architecture.

Where Pith is reading between the lines

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

  • The lateral-root-as-object strategy should transfer to other self-occluding, texture-poor branching structures — tree canopies, vasculature, neurons — wherever branches can be defined as detections.
  • The clean-mesh testbed with known poses is the main gap between these numbers and field performance; the decisive next test is unknown-pose photography of real soil-grown roots against CT or MRI ground truth.
  • The growth-simulation rule for connecting lateral roots assumes downward, tip-first growth; root systems with horizontal or upward-growing branches would need a learned connectivity head.
  • If pose-noise robustness holds, the pipeline could run on video from a single moving camera, making 3D root phenotyping possible in the field rather than the lab.
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 / 7 minor

Summary. The manuscript proposes a pipeline for extracting 3D plant root skeletons from 3–10 multi-angle images. The method detects lateral roots with a YOLOv8-based detector that also predicts root keypoints, matches lateral roots across views using LightGlue keypoints and a voting scheme, triangulates the start/end points with weighted reprojection-error optimization, connects lateral roots to a main root using a growth-simulating row-propagation algorithm, and refines the skeleton and camera poses with a differentiable skeleton bundle adjustment (SBA) layer. Experiments on a custom dataset of 400 sweet-potato root mesh models, with rendered projection images and 40 test roots, report 2D AP/AR of 0.81/0.69 and 3D AP/AR of 0.77/0.58, outperforming AdaBins, MIM, and Depth Anything followed by Pc-Skeletor skeletonization. The paper claims to be the first to extract 3D root skeletons from images and to support automated breeding robots.

Significance. If the method works on real imagery, the contribution is potentially valuable for root phenotyping and breeding robotics. The modular design—detection, matching, triangulation, and SBA—is reasonable, and the idea of using skeletal-angle consistency for self-supervised bundle adjustment is interesting. The underlying triangulation and bundle-adjustment mathematics are standard and internally coherent. However, the only quantitative evaluation is performed on projection images rendered from manually denoised 3D meshes with known camera poses. This removes the key difficulties of real root imaging: occlusions, soil clutter, near-textureless surfaces, and unknown pose estimation. The reported AP/AR therefore measure performance on a clean synthetic-like proxy, not on the claimed application domain. The central claim is plausible but not yet supported; the paper needs either real-image validation or a substantial reduction in the scope of its claims. No code or dataset is released, and the evaluation protocol is not fully specified.

major comments (4)
  1. [Section V.A, Table III] The load-bearing claim in the abstract and contribution 3—that the method extracts 3D root skeletons 'from a few images'—is evaluated only on projection images rendered from manually denoised 3D mesh models. The virtual camera positions are known when rendering, so the triangulation in Section IV.A and the SBA initialization do not face the pose-estimation problem that real multi-view root images would present. Moreover, rendered meshes provide shading and edge cues that real near-textureless roots lack, and the images are free of soil occlusion and clutter. Consequently, the AP/AR values in Table III do not support the stated real-world application. Please add experiments on real multi-view root images (including occlusions and estimated camera poses) or explicitly restrict the claims to clean rendered models.
  2. [Section IV.C, Eqs. (10)-(11)] The main-root connection algorithm assumes that in each image row the foreground is a single continuous interval [x1, x2]. This is true for the clean renderings used in Section V.A but is unlikely to hold for real root images with soil, clutter, crossed branches, or fragmented foreground regions. The mode-based propagation rule m = argmax_v |{M_(y-1)j = v}| depends critically on this continuity assumption. Since main-root skeleton extraction is an integral part of the method's central contribution, the algorithm should be tested on occluded or fragmented foregrounds, or modified to handle multiple intervals per row.
  3. [Tables I-III, Section V.C] The precision and recall metrics for lateral-root detection and 3D skeleton extraction are not formally defined. It is unclear how predicted lateral roots or 3D skeletons are matched to ground-truth roots: is an IoU threshold, a point-to-mesh distance, or a graph-matching criterion used? Without this protocol, the reported numbers (e.g., 3D AP 0.77 and AR 0.58) cannot be independently interpreted or reproduced. Please specify the matching criterion, including thresholds and how multiple predictions per root are handled.
  4. [Section IV.B] The skeleton-angle self-supervision for the SBA layer is underspecified. The text says the loss is 'the difference between the original angle and the new angle after projection,' but it does not define what the 'original angle' is, when it is computed, or how 'new angle after projection' is obtained. If the original angle is measured from the current 3D skeleton estimate, the loss may act only as a regularizer rather than as a ground-truth-free self-supervision signal. Additionally, no ablation study isolates the contribution of the SBA layer to the final AP/AR, so the benefit of Eq. (9) is not demonstrated. Please provide a formal definition of the angle loss and add an SBA on/off comparison.
minor comments (7)
  1. [Fig. 1 caption] Typo: 'keleton extraction netwpork' should be 'skeleton extraction network'.
  2. [Eq. (2)] The y-coordinate keypoint scaling uses Aw_s (anchor width) for both b_y and b_x; presumably the y-coordinate should be scaled by Ah_s. Please correct.
  3. [Eq. (8)] The notation is inconsistent: Eq. (8) writes π(T_i, p_i) but the subscript in eg_i,j is j; it should likely be π(T_i, p_j). Please fix the index.
  4. [Section IV.A] The sentence 'For lateral roots matched only in the second image' appears to mean the third image in the context of the pairwise triangulation; please clarify.
  5. [References] Reference [2] is cited in the COLMAP discussion in Section II.B, but the work is about poultry farms and appears unrelated. Please verify the citation.
  6. [Table III] The comparison baselines are monocular depth estimators (AdaBins, MIM, Depth Anything) followed by Pc-Skeletor; classical multi-view stereo such as COLMAP is mentioned in the related work but not compared. Adding such a baseline would make the comparison more complete.
  7. [General] No code, trained models, or the custom dataset are released, which limits reproducibility; consider providing these or a detailed protocol.

Circularity Check

0 steps flagged

No circularity: the reconstruction pipeline does not use ground-truth skeletons as inputs; the main limitation is synthetic-domain evaluation rather than self-referential derivation.

full rationale

The claimed derivation chain is input-to-output and uses ground truth only for final evaluation. Lateral-root detection is trained with objectness/box/keypoint losses (Eqs. 3-6) on the authors' rendered dataset; matching is done by LightGlue keypoint voting; 3D coordinates are obtained by weighted triangulation from multiple views; the SBA layer (Eqs. 7-9) optimizes camera poses and skeleton points by reprojection error, with the skeletal-angle loss acting as a self-supervised geometric regularizer rather than as a fitted copy of the annotation. The ground-truth meshes are never fed into the network at inference, and the test roots are separate from the training roots. The many self-citations in the related work (e.g., [15], [18], [19], [20], [33]-[35]) are contextual and do not carry the burden of the proposed derivation. The principal weakness is external validity, not circularity: projection images are rendered from manually denoised mesh models with known camera poses, so the testbed omits soil occlusion, unknown poses, and real-texture challenges. That is a domain-mismatch/generalization concern; it does not make the predicted skeleton equivalent to the input or to the ground truth by construction. No equation or fitted parameter in the paper reduces the claimed 'prediction' to its own input.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 0 invented entities

The central result does not introduce new physical entities. It does depend on a set of domain assumptions about root shape, growth direction, camera geometry, and the validity of synthetic projection images as a stand-in for real photographs.

free parameters (5)
  • Matching Threshold = not reported
    Minimum number of matched keypoints required to accept a lateral-root bounding-box pair; set by hand in Section III.B.
  • Grid weighting factor omega_s = not reported
    Per-scale loss weights in Eq. 3; values are not given.
  • Loss weights lambda_obj, lambda_box, lambda_kps = not reported
    Trade-off weights in Eq. 6; values are not given.
  • Lateral-root removal threshold = not reported
    Threshold distance used in Section IV.A to remove reprojected lateral roots that fall into background or far from the root.
  • Minimum connection count = 2
    In Section IV.C, only main-root connections appearing at least twice are retained; chosen by hand.
axioms (5)
  • domain assumption Multi-view images have known camera poses or poses that SBA can refine from a good initialization
    Triangulation in Section IV.A and the SBANet input T = {Ti} in Section IV.B require this; no real-image calibration or pose-estimation procedure is described.
  • domain assumption Projection images generated from manually denoised 3D meshes are representative of real root images
    The dataset in Section V.A is built by generating 3 to 10 projection images from mesh models, which removes soil, lighting, and unknown camera geometry.
  • domain assumption Each lateral root can be approximated by a straight skeleton between its detected start and end keypoints
    Sections III.A and IV.A connect only start and end points and do not model within-branch curvature.
  • domain assumption Main roots grow downward from the top of the image, so row-wise propagation connects lateral roots correctly
    Section IV.C simulates root growth by updating rows from smallest to largest y value.
  • domain assumption Angles between roots are invariant in 3D and measurable in projections, providing valid self-supervision
    Section IV.B uses skeletal angle differences as the training loss for SBANet.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of 3D Plant Root Skeleton Detection and Extraction." pith.science (2026). https://pith.science/paper/OODAGART

@misc{pith2026250808094,
  author       = {Pith},
  title        = {Pith review of: 3D Plant Root Skeleton Detection and Extraction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OODAGART}},
  note         = {Machine review of arXiv:2508.08094}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Plant roots typically exhibit a highly complex and dense architecture, incorporating numerous slender lateral roots and branches, which significantly hinders the precise capture and modeling of the entire root system. Additionally, roots often lack sufficient texture and color information, making it difficult to identify and track root traits using visual methods. Previous research on roots has been largely confined to 2D studies; however, exploring the 3D architecture of roots is crucial in botany. Since roots grow in real 3D space, 3D phenotypic information is more critical for studying genetic traits and their impact on root development. We have introduced a 3D root skeleton extraction method that efficiently derives the 3D architecture of plant roots from a few images. This method includes the detection and matching of lateral roots, triangulation to extract the skeletal structure of lateral roots, and the integration of lateral and primary roots. We developed a highly complex root dataset and tested our method on it. The extracted 3D root skeletons showed considerable similarity to the ground truth, validating the effectiveness of the model. This method can play a significant role in automated breeding robots. Through precise 3D root structure analysis, breeding robots can better identify plant phenotypic traits, especially root structure and growth patterns, helping practitioners select seeds with superior root systems. This automated approach not only improves breeding efficiency but also reduces manual intervention, making the breeding process more intelligent and efficient, thus advancing modern agriculture.

Figures

Figures reproduced from arXiv: 2508.08094 by Ge Jin, Guoyu Lu, Jiakai Lin, Jinchang Zhang, Tianming Liu, Wenzhan Song.

Figure 1
Figure 1. Figure 1: Overview of our root keleton extraction netwpork structure, which includes a detection network for identifying the positions and key points of [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The Structure of Skeleton Bundle Adjustment (SBA) Net [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: 3D skeleton results: (a), (b) is different views, (c) is AdaBins [1], (d) is MIM [29], (e) is Depth Anything [32], (f), (g), (h) showcase the results [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: 3D skeleton results: (a) is input. (b) is Ground Truth. (c) is Plant [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

36 extracted references · 34 canonical work pages · 1 internal anchor

  1. [1]

    Adabins: Depth estimation using adaptive bins

    Shariq Farooq Bhat, Ibraheem Alhashim, and Peter Wonka. Adabins: Depth estimation using adaptive bins. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 4009– 4018, 2021

    work page 2021

  2. [2]

    Developing edge ai computer vision for smart poultry farms using deep learning and hpc

    Stevan Cakic, Tomo Popovic, Srdjan Krco, Daliborka Nedic, Dejan Babic, and Ivan Jovovic. Developing edge ai computer vision for smart poultry farms using deep learning and hpc. Sensors, 23(6):3002, 2023

    work page 2023

  3. [3]

    Deep-based ingredient recogni- tion for cooking recipe retrieval

    Jingjing Chen and Chong-Wah Ngo. Deep-based ingredient recogni- tion for cooking recipe retrieval. In Proceedings of the 24th ACM international conference on Multimedia , pages 32–41, 2016

    work page 2016

  4. [4]

    Cross-modal recipe retrieval with rich food attributes

    Jing-jing Chen, Chong-Wah Ngo, and Tat-Seng Chua. Cross-modal recipe retrieval with rich food attributes. In Proceedings of the 25th ACM international conference on Multimedia , pages 1771–1779, 2017

    work page 2017

  5. [5]

    3d-r2n2: A unified approach for single and multi- view 3d object reconstruction

    Christopher B Choy, Danfei Xu, JunYoung Gwak, Kevin Chen, and Silvio Savarese. 3d-r2n2: A unified approach for single and multi- view 3d object reconstruction. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part VIII 14 , pages 628–644, 2016

    work page 2016

  6. [6]

    Su- perpoint: Self-supervised interest point detection and description

    Daniel DeTone, Tomasz Malisiewicz, and Andrew Rabinovich. Su- perpoint: Self-supervised interest point detection and description. In Proceedings of the IEEE conference on computer vision and pattern recognition workshops, pages 224–236, 2018

    work page 2018

  7. [7]

    Investigation of different skeleton features for cnn-based 3d action recognition

    Zewei Ding, Pichao Wang, Philip O Ogunbona, and Wanqing Li. Investigation of different skeleton features for cnn-based 3d action recognition. In IEEE International conference on multimedia & expo workshops (ICMEW), pages 617–622, 2017

    work page 2017

  8. [8]

    A point set gener- ation network for 3d object reconstruction from a single image

    Haoqiang Fan, Hao Su, and Leonidas J Guibas. A point set gener- ation network for 3d object reconstruction from a single image. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 605–613, 2017

    work page 2017

  9. [9]

    Visual simultaneous localization and mapping: a survey

    Jorge Fuentes-Pacheco, Jos ´e Ruiz-Ascencio, and Juan Manuel Rend´on-Mancha. Visual simultaneous localization and mapping: a survey. Artificial intelligence review, 43:55–81, 2015

    work page 2015

  10. [10]

    Deepmvs: Learning multi-view stereopsis

    Po-Han Huang, Kevin Matzen, Johannes Kopf, Narendra Ahuja, and Jia-Bin Huang. Deepmvs: Learning multi-view stereopsis. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2821–2830, 2018

    work page 2018

  11. [11]

    Surfacenet: An end-to-end 3d neural network for multiview stereopsis

    Mengqi Ji, Juergen Gall, Haitian Zheng, Yebin Liu, and Lu Fang. Surfacenet: An end-to-end 3d neural network for multiview stereopsis. In Proceedings of the IEEE international conference on computer vision, pages 2307–2315, 2017

    work page 2017

  12. [12]

    Ultralytics YOLO, Jan

    Glenn Jocher, Ayush Chaurasia, and Jing Qiu. Ultralytics YOLO, Jan. 2023

    work page 2023

  13. [13]

    Root system architecture and abiotic stress tolerance: current knowledge in root and tuber crops

    M Awais Khan, Dorcus C Gemenet, and Arthur Villordon. Root system architecture and abiotic stress tolerance: current knowledge in root and tuber crops. Frontiers in Plant Science , 7:209303, 2016

    work page 2016

  14. [14]

    Skeleton based action recognition using translation-scale invariant image mapping and multi-scale deep cnn

    Bo Li, Yuchao Dai, Xuelian Cheng, Huahui Chen, Yi Lin, and Mingyi He. Skeleton based action recognition using translation-scale invariant image mapping and multi-scale deep cnn. In IEEE International Conference on Multimedia & Expo Workshops (ICMEW) , pages 601– 604, 2017

    work page 2017

  15. [15]

    Keypoint Detection and Description for Raw Bayer Images

    Jiakai Lin, Jinchang Zhang, and Guoyu Lu. Keypoint detection and description for raw bayer images. arXiv preprint arXiv:2503.08673 , 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  16. [16]

    Light- glue: Local feature matching at light speed

    Philipp Lindenberger, Paul-Edouard Sarlin, and Marc Pollefeys. Light- glue: Local feature matching at light speed. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 17627–17638, 2023

    work page 2023

  17. [17]

    A simple pooling-based design for real-time salient object detection

    Jiang-Jiang Liu, Qibin Hou, Ming-Ming Cheng, Jiashi Feng, and Jianmin Jiang. A simple pooling-based design for real-time salient object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 3917–3926, 2019

    work page 2019

  18. [18]

    Non-destructive 3d root structure modeling

    Guoyu Lu. Non-destructive 3d root structure modeling. In IEEE International Conference on Robotics and Automation (ICRA) , 2025

    work page 2025

  19. [19]

    3d plant root system reconstruction based on fusion of deep structure-from-motion and imu

    Yawen Lu, Yuxing Wang, Zhanjie Chen, Awais Khan, Carl Salvag- gio, and Guoyu Lu. 3d plant root system reconstruction based on fusion of deep structure-from-motion and imu. Multimedia tools and applications, 80(11):17315–17331, 2021

    work page 2021

  20. [20]

    Simultaneous direct depth estimation and synthesis stereo for single image plant root reconstruction

    Yawen Lu, Yuxing Wang, Devarth Parikh, Awais Khan, and Guoyu Lu. Simultaneous direct depth estimation and synthesis stereo for single image plant root reconstruction. IEEE Transactions on Image Processing, 30:4883–4893, 2021

    work page 2021

  21. [21]

    Rethinking keypoint representations: Modeling keypoints and poses as objects for multi-person human pose estimation

    William McNally, Kanav Vats, Alexander Wong, and John McPhee. Rethinking keypoint representations: Modeling keypoints and poses as objects for multi-person human pose estimation. In European Conference on Computer Vision , pages 37–54, 2022

    work page 2022

  22. [22]

    Cherrypicker: Semantic skeletonization and topological reconstruction of cherry trees

    Lukas Meyer, Andreas Gilson, Oliver Scholz, and Marc Stamminger. Cherrypicker: Semantic skeletonization and topological reconstruction of cherry trees. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 6243–6252, 2023

    work page 2023

  23. [23]

    Phenotyping of architecture traits of loblolly pine trees using stereo machine vision and deep learning: Stem diameter, branch angle, and branch diameter

    Nariman Niknejad, Rafael Bidese-Puhl, Yin Bao, Kitt G Payn, and Jingyi Zheng. Phenotyping of architecture traits of loblolly pine trees using stereo machine vision and deep learning: Stem diameter, branch angle, and branch diameter. Computers and Electronics in Agriculture, 211:107999, 2023

    work page 2023

  24. [24]

    A survey of structure from motion*

    Onur ¨Ozyes ¸il, Vladislav V oroninski, Ronen Basri, and Amit Singer. A survey of structure from motion*. Acta Numerica, 26:305–364, 2017

    work page 2017

  25. [25]

    Regulation of plant root system architecture: implications for crop advancement

    Eric D Rogers and Philip N Benfey. Regulation of plant root system architecture: implications for crop advancement. Current Opinion in Biotechnology, 32:93–98, 2015

    work page 2015

  26. [26]

    Ba-net: Dense bundle adjustment network

    Chengzhou Tang and Ping Tan. Ba-net: Dense bundle adjustment network. arXiv preprint arXiv:1806.04807 , 2018

    Pith/arXiv arXiv 2018

  27. [27]

    Pixel2mesh: Generating 3d mesh models from single rgb images

    Nanyang Wang, Yinda Zhang, Zhuwen Li, Yanwei Fu, Wei Liu, and Yu-Gang Jiang. Pixel2mesh: Generating 3d mesh models from single rgb images. In Proceedings of the European conference on computer vision (ECCV), pages 52–67, 2018

    work page 2018

  28. [28]

    Action recognition based on joint trajectory maps using convolutional neural networks

    Pichao Wang, Zhaoyang Li, Yonghong Hou, and Wanqing Li. Action recognition based on joint trajectory maps using convolutional neural networks. In Proceedings of the 24th ACM international conference on Multimedia, pages 102–106, 2016

    work page 2016

  29. [29]

    Revealing the dark secrets of masked image modeling

    Zhenda Xie, Zigang Geng, Jingcheng Hu, Zheng Zhang, Han Hu, and Yue Cao. Revealing the dark secrets of masked image modeling. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14475–14485, 2023

    work page 2023

  30. [30]

    Ensemble one-dimensional convolution neural networks for skeleton-based action recognition

    Yangyang Xu, Jun Cheng, Lei Wang, Haiying Xia, Feng Liu, and Dapeng Tao. Ensemble one-dimensional convolution neural networks for skeleton-based action recognition. IEEE Signal Processing Letters, 25(7):1044–1048, 2018

    work page 2018

  31. [31]

    3d object reconstruction from a single depth view with adversarial learning

    Bo Yang, Hongkai Wen, Sen Wang, Ronald Clark, Andrew Markham, and Niki Trigoni. 3d object reconstruction from a single depth view with adversarial learning. In Proceedings of the IEEE international conference on computer vision workshops , pages 679–688, 2017

    work page 2017

  32. [32]

    Depth anything: Unleashing the power of large-scale unlabeled data

    Lihe Yang, Bingyi Kang, Zilong Huang, Xiaogang Xu, Jiashi Feng, and Hengshuang Zhao. Depth anything: Unleashing the power of large-scale unlabeled data. arXiv preprint arXiv:2401.10891 , 2024

    Pith/arXiv arXiv 2024

  33. [33]

    Underground mapping and localiza- tion based on ground-penetrating radar

    Jinchang Zhang and Guoyu Lu. Underground mapping and localiza- tion based on ground-penetrating radar. In Proceedings of the Asian Conference on Computer Vision , pages 2018–2033, 2024

    work page 2018

  34. [34]

    Vision-language embodiment for monocular depth estimation

    Jinchang Zhang and Guoyu Lu. Vision-language embodiment for monocular depth estimation. In Proceedings of the Computer Vision and Pattern Recognition Conference , pages 29479–29489, 2025

    work page 2025

  35. [35]

    Embodiment: Self-supervised depth estimation based on camera models

    Jinchang Zhang, Praveen Kumar Reddy, Xue-Iuan Wong, Yiannis Aloimonos, and Guoyu Lu. Embodiment: Self-supervised depth estimation based on camera models. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 7809– 7816, 2024

    work page 2024

  36. [36]

    Distance-iou loss: Faster and better learning for bounding box regression

    Zhaohui Zheng, Ping Wang, Wei Liu, Jinze Li, Rongguang Ye, and Dongwei Ren. Distance-iou loss: Faster and better learning for bounding box regression. In Proceedings of the AAAI conference on artificial intelligence, volume 34, pages 12993–13000, 2020

    work page 2020

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