Pith. sign in

REVIEW 4 major objections 6 minor 2 cited by

FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read FruitNeRF++ claims that fruit counting can be done without fruit-specific templates by learning per-fruit instance embeddings in a neural radiance field.

desk verdict A solid extension of FruitNeRF that overstates its fruit-agnostic claim; the core idea is sound, but the evaluation doesn't support end-to-end generalization. read the letter →

arxiv 2505.19863 v1 pith:SW5ZKR3G submitted 2025-05-26 cs.CV cs.LG

Lukas Meyer , Andrei-Timotei Ardelean , Tim Weyrich , Marc Stamminger This is my paper · ORCID
classification cs.CVcs.LG
keywords fruitcountingneuralradiancefieldscontrastivelearninginstancesegmentationpointcloudclusteringshape-agnosticagriculturalcomputervisionHDBSCAN
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

FruitNeRF++ aims to make orchard fruit counting independent of fruit type. Its central claim is that a neural radiance field can learn a per-fruit identity field, trained by contrastive learning on 2D instance masks predicted by vision foundation models, so that counting reduces to clustering a point cloud whose points carry both fruit semantics and instance embeddings. No fruit-specific shape templates are needed, which addresses the main limitation of the prior FruitNeRF method. The paper shows the approach on six synthetic fruit types and a real apple dataset, and argues the same machinery applies to arbitrary objects.

What carries the argument

The central object is the neural instance field, a function $F_i$ mapping a 3D point to a $D$-dimensional identity embedding, rendered volumetrically like color and semantics. It is trained with a contrastive loss in which each sampled pixel is attracted to the mean prototype of its own fruit mask and repelled from all other fruit prototypes, with a pixel sampler that mines nearby fruits as hard negatives. A cascaded training scheme first fits density and RGB, then adds the semantic fruit field, then freezes both and trains only the instance field. The final counting step partitions the fruit point cloud into spatial chunks with $k$-Means and clusters each chunk with HDBSCAN using distance $d = \lambda_c d_c + \lambda_e d_e$ with $\lambda_c = 1$ and $\lambda_e = 5$ in the real-world setting.

What would settle it

On a synthetic scene with ground-truth instance IDs, render leaves that partially occlude one fruit and feed the resulting automatic masks to the pipeline; if the occluded fruit is counted as two while all unobstructed fruits are counted correctly, the load-bearing assumption of mask consistency is falsified for that setting.

Watch

Extended reading notes

Core claim

The paper's central discovery is that instance masks, even when they are inconsistent across views, can be fused into a 3D-consistent instance field by contrastive learning: the same fruit's pixels are attracted to a shared prototype embedding while different fruits are repelled. Rendering this field volumetrically and sampling it yields a fruit point cloud with D-dimensional identity features, and clustering with HDBSCAN under a cosine-plus-Euclidean distance gives the count. The authors report that this removes the need for per-fruit templates and that the pipeline, with identical parameters, counts apples, plums, lemons, pears, peaches, and mangoes, plus 99 of 100 common household objects in a separate scene. On the real FUJI apple dataset the F1-score is 0.765 with automatically predicted masks.

Load-bearing premise

The instance masks from SAM/Detic must assign a consistent identity to each whole fruit across all views; when a leaf or branch splits a fruit into two masks, the contrastive loss pulls the two halves' embeddings apart and the count doubles.

Editorial extensions

If this is right

  • The same pipeline, with no fruit-specific parameters, counts apples, plums, lemons, pears, peaches, and mangoes in synthetic scenes, and the authors state the method is agnostic to fruit type in both the neural field and the clustering.
  • Users no longer need to construct or tune fruit-specific templates or cascaded clustering hyper-parameters, making the method easier to control than FruitNeRF.
  • Because only RGB images and automatic instance masks are required, the approach extends to arbitrary object types; the authors demonstrate this by detecting 99 of 100 objects in a Messy Rooms household scene.
  • The counting result depends on the quality of the mask predictions: average F1 over synthetic fruits is 0.925 with ground-truth masks, 0.832 with Grounded-SAM masks, and 0.776 with Detic masks.
  • The method's speed makes it unsuitable for real-time use: training the FUJI scene takes about 8 hours on an A5000 GPU, with the instance field converging slowly.

Reading between the lines

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

  • Because the failure mode is mask splitting, a direct extension is to add a post-hoc merging step: cluster instance embeddings first and then split clusters whose embedding shows a bimodal structure, which could recover fruits split by occluders.
  • The contrastive prototype loss could be combined with temporal or geometric priors (e.g., fruit centers from the density field) to regularize embeddings for clustered fruits like bananas or berries, which the paper lists as an open challenge.
  • If the method transfers to arbitrary objects as the Messy Rooms result suggests, it could be used to count other small, partially occluded clusters in the wild, such as flowers, berries on bushes, or manufactured parts in bins, with only text-prompted masks.
  • The reported dependence on the Euclidean weight $\lambda_e$ hints that the instance embeddings alone carry less identity signal than the geometry in sparse scenes; on denser or more occluded scenes, higher embedding dimensions and temperature tuning may become critical.
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

4 major / 6 minor

Summary. FruitNeRF++ proposes a fruit-counting pipeline that combines a neural radiance field with a contrastively trained neural instance field. The method takes 2D instance masks from foundation models (Grounded-SAM, Detic) and lifts them into 3D via contrastive learning, then produces a fruit count by clustering instance-embedded fruit points with HDBSCAN using a Euclidean-plus-cosine distance. The paper evaluates on a synthetic multi-fruit dataset (apple, plum, lemon, pear, peach, mango) and on the real FUJI apple dataset, reporting F1-scores for different mask sources and comparing with FruitNeRF and Gené-Mola et al. The central claim is that the method is fully fruit-agnostic in both the neural field and the clustering stages.

Significance. If the method delivered on its claims, it would be a valuable step toward practical multi-fruit counting without fruit-specific shape templates. The contrastive instance-field formulation with local hard negatives for small objects is a plausible and potentially useful contribution, and the paper ships a project page and promises code, which increases reproducibility. However, the strength of the claims currently exceeds the evidence: the 'fully fruit-agnostic' conclusion is contradicted by the authors' own fruit-specific prompt adjustments and by the large mask-quality-dependent performance gaps; the hyperparameter protocol is inconsistent; and the real-world comparison does not support the abstract's 'compares favorably' wording. The core idea is promising, but the evaluation and claims need substantial rework before the manuscript meets the standard of a journal publication.

major comments (4)
  1. [Sec. IV-A, Sec. VII, Table II] The claim in Sec. VII that the approach is 'fully agnostic to the type of fruit, regarding both the neural radiance field and the clustering' is contradicted by the method description in Sec. IV-A, where the authors state that for plum and mango they added 'apple' to the Grounded-SAM prompt because the fruit-specific prompt alone produced poor masks. This is an explicit fruit-specific intervention in the pipeline. Moreover, Table II shows that with SAM-generated masks, plum recall drops to 0.341 and mango recall to 0.392 (from 0.759 and 0.557 with GT masks), while apple, lemon, pear, and peach remain near 0.94–1.0. This pattern tracks the segmentation IoU in Table I (plum 0.382, mango 0.588) rather than any property of the NeRF or clustering stage. The end-to-end method is therefore not fruit-agnostic; it is a shape-agnostic clustering scheme whose counting accuracy is dominated by a fruit-dependent segmentation module. Please rephrase the central claim to describe the NeRF and clustering stages as shape-agnostic, and add a quantitative analysis of how mask quality propagates to counting error.
  2. [Sec. IV-E, Sec. V, Fig. 5] There is a direct inconsistency in the hyperparameters. Sec. IV-E states that the Euclidean distance weight in Eq. (9) is set to lambda_e = 5, while Sec. V begins by saying 'We used the same parameters for all experiments' and lists lambda_e = 1. Later in Sec. V, the FUJI experiment is reported with tau = 0.35 and lambda_e = 5, again different from the synthetic experiments. Given that Fig. 5 shows the results are sensitive to these parameters, this inconsistency suggests that hyperparameters were selected per dataset without a held-out validation split. The paper reports no error bars and no validation protocol, making it impossible to assess whether the reported numbers generalize. Please clarify the exact hyperparameter settings used for each reported result, justify any differences, and provide a validation procedure (e.g., tune on a subset, report variance across runs or splits).
  3. [Sec. V, Abstract] The abstract claims the method 'compares favorably to other state-of-the-art methods,' but the only real-world comparison reported in Sec. V gives FruitNeRF++ an F1-score of 0.765 on the FUJI dataset, whereas Gené-Mola et al. [2] achieve 0.881. The numerical result does not support the phrasing 'compares favorably.' The authors attribute the gap to 'noisy poses and manual registration of both sides,' but no ablation or quantitative analysis is provided to substantiate this explanation. Please either temper the claim to a direct comparison of the reported numbers, or provide supporting experiments that isolate the effect of pose noise (e.g., using refined poses or a subset of images with more reliable registration).
  4. [Table II, Sec. IV-C] The GT-mask experiments in Table II are a partially circular sanity check: the instance field is trained using ground-truth instance masks (Sec. IV-C.4) and then the counting result is evaluated against the same ground-truth fruit centers. Since the model has been given the correct instance identities during training, the clustering step is essentially verifying that the learned embeddings separate under the same labels used for supervision. These experiments are therefore useful for debugging the clustering and the field, but they cannot be taken as evidence of generalization. The paper should clearly frame the GT-mask results as a sanity check and base claims of practical performance on the SAM/Detic mask experiments, which are the ones that reflect the intended use case.
minor comments (6)
  1. [Sec. II-B] The word 'constrastive' in the phrase 'a constrastive objective' is a typo and should be 'contrastive.'
  2. [Sec. IV-C.4] The phrase 'prevent both gradients to propagate trough the density field' contains a typo: 'trough' should be 'through.'
  3. [Sec. V] The sentence 'The results of our counting algorithm compared to Gené-Mola et al. [2] can be attributed to the noisy poses and manual registration of both sides' is grammatically awkward and the assertion is not backed by evidence; please rephrase and provide support or remove the attribution.
  4. [Sec. VII] The phrase 'Messy Rooms dataset dataset' has a duplicated word 'dataset.'
  5. [Table I] The caption states 'The best performing segmentation model is highlighted in green,' but the table as rendered does not appear to contain any green highlighting, and the cell 'see caption' is unclear; please ensure the table matches the caption or revise the caption.
  6. [Sec. VII] The experiment on the Messy Rooms dataset is only mentioned in the conclusion with the result 'we detected 99 objects,' with no description of the setup, the number of objects, the metric, or the parameters used; please move this to the evaluation section or remove it.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central fruit count is evaluated against independent ground-truth counts using predicted masks; the GT-mask table and prototype-based contrastive target are internal consistency checks, not load-bearing predictions.

full rationale

The paper's claimed derivation chain is: input RGB images plus semantic and instance masks (GT or from SAM/Detic); train density, appearance, semantic, and instance fields; export a fruit point cloud with instance embeddings; cluster with HDBSCAN using a Euclidean-plus-cosine metric; the cluster count is the fruit count. The main evaluation uses masks predicted by Grounded-SAM/Detic and compares the resulting count against independent ground-truth fruit counts, so the central claim is externally falsifiable rather than forced by construction. The GT-mask experiments in Table II train the instance field on ground-truth instance identities and then evaluate counting against the same ground-truth identities; this is a controlled sanity check of the readout pipeline, not the paper's central claim, and it is not a reduction by construction because clustering can and does fail (e.g., mango recall 0.557 with GT masks). The contrastive loss in Eq. 7 uses a fruit-prototype feature vector that is the average of the current features, making the positive target self-referential in the training objective, but this is a standard contrastive formulation and does not by itself force the final count; the counting result is still determined by the subsequent clustering and is benchmarked externally. The self-citation to FruitNeRF [6] supplies the base architecture and dataset, but the instance-field mechanism and its evaluation are new and independent. The fruit-specific prompt additions for plum and mango, and the Fuji comparison where FruitNeRF++ (F1=0.765) trails Gené-Mola et al. (F1=0.881), are overclaims or consistency issues rather than circular reasoning. Overall, no load-bearing step reduces to its own inputs by the paper's equations or by a self-citation chain.

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

The central claim rests on several tuned hyperparameters and domain assumptions. The most important missing pieces are the unspecified sampling counts and the lack of held-out validation for the chosen hyperparameters.

free parameters (5)
  • Instance embedding dimension D = 32 (synthetic and FUJI)
    Swept from 1 to 256 in Fig. 5 (left); final value chosen by best F1 on the evaluation set.
  • Contrastive temperature tau = 0.2 (synthetic), 0.35 (FUJI)
    Swept in Fig. 5 (center); different values used per dataset; no held-out validation.
  • Clustering weight lambda_e = 1 (synthetic), 5 (FUJI)
    Swept in Fig. 5 (right) and Section V; value selected per dataset to maximize F1.
  • Spatial partitions S = 40 (FUJI)
    Arbitrarily selected to reduce point cloud size in clustering; no analysis of sensitivity for synthetic data.
  • Sampling counts G, L, P = not specified
    Number of groups, fruits per group, and pixels per fruit in the contrastive sampler; no values given, so exact reproduction is impossible.
assumptions (4)
  • domain assumption Instance masks from foundation models provide reliable per-fruit identity across views
    Section IV-B and VI; the method relies on SAM/Detic masks; the authors note masks can split occluded fruits.
  • domain assumption Camera poses recovered by COLMAP are accurate enough for the instance field to converge
    Section V and VI; 'noisy data significantly decreases the correctness of the implicit field... mainly due to inaccurate poses'.
  • domain assumption Fruits are spatially separable such that Euclidean distance contributes to clustering
    Eq. 9 uses Euclidean distance with weight 5; clustered fruits or heavy occlusion would break this.
  • standard math Standard NeRF volumetric rendering integrates fields correctly
    Used in Equations 1-3; standard background from Mildenhall et al.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields." pith.science (2026). https://pith.science/paper/SW5ZKR3G

@misc{pith2026250519863,
  author       = {Pith},
  title        = {Pith review of: FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SW5ZKR3G}},
  note         = {Machine review of arXiv:2505.19863}
}
read the original abstract

We introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards. Our work is based on FruitNeRF, which employs a neural semantic field combined with a fruit-specific clustering approach. The requirement for adaptation for each fruit type limits the applicability of the method, and makes it difficult to use in practice. To lift this limitation, we design a shape-agnostic multi-fruit counting framework, that complements the RGB and semantic data with instance masks predicted by a vision foundation model. The masks are used to encode the identity of each fruit as instance embeddings into a neural instance field. By volumetrically sampling the neural fields, we extract a point cloud embedded with the instance features, which can be clustered in a fruit-agnostic manner to obtain the fruit count. We evaluate our approach using a synthetic dataset containing apples, plums, lemons, pears, peaches, and mangoes, as well as a real-world benchmark apple dataset. Our results demonstrate that FruitNeRF++ is easier to control and compares favorably to other state-of-the-art methods.

Figures

Figures reproduced from arXiv: 2505.19863 by the authors.

Figure 1
Figure 1. Rendering of RGB, semantic and instance images. For visualization of the results visit the project page: [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Pipeline of FruitNeRF++. For the images we recover both intrinsic and extrinsic camera parameters. We then extract semantic and instance masks for arbitrary fruit types using SAM [39] and Detic [41]. The data are used to train a neural radiance field with an neural appearance, semantic (fruit) and instance field. By clustering the combination of a fruit and instance point cloud we obtain a precise fruit count. 1 − e… view at source ↗
Figure 3
Figure 3. Overview of the FruitNeRF++ architecture, split up into four different components: density field, appearance field, fruit field, and instance field. The density field encodes the volume density 𝜎, the appearance field the color RGB, the Fruit Field the semantic information about the fruit in space, and the instance field a feature vector i encoding information about the instance group of a point in space. The dashed… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: In Fig. (a) we visualize the concept of local and global [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Counting result on variation of the embedding size [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Towards Integrating Multi-Spectral Imaging with Gaussian Splatting

    cs.CV 2025-08 conditional novelty 5.0 of 10

    Jointly optimizing RGB and four additional spectral bands in one 3D Gaussian Splatting model, after an RGB-only warm-up and with spectrum-aware densification, outperforms per-band models and slightly improves RGB via ...

  2. CountingFruit: Language-Guided 3D Fruit Counting with Semantic Gaussian Splatting

    cs.CV 2025-06 conditional novelty 4.0 of 10

    FruitLangGS combines language-guided Gaussian splatting with prompt-based filtering and clustering to count apples in 3D orchard scenes, reporting up to 99.7% recall.

Reference graph

Works this paper leans on

52 extracted references · 51 canonical work pages · cited by 2 Pith papers

  1. [2]

    Fruit detection and 3D location using instance segmentation neural networks and structure-from-motion photogrammetry,

    J. Gené-Mola, R. Sanz-Cortiella, J. R. Rosell-Polo, J. R. Morros, J. Ruiz- Hidalgo, V. Vilaplana and E. Gregorio, "Fruit detection and 3D location using instance segmentation neural networks and structure-from-motion photogrammetry,"Computers and Electronics in Agriculture, 2020

    work page 2020

  2. [1]

    Fruit sizing using AI: A review of methods and challenges,

    J. C. Miranda, J. Gené-Mola, M. Zude-Sasse, N. Tsoulias, A. Escolà, J. Arnó, J. R. Rosell-Polo, R. Sanz-Cortiella, J. A. Martínez-Casasnovas and E. Gregorio, "Fruit sizing using AI: A review of methods and challenges,"Postharvest Biology and Technology, 2023

    work page 2023

  3. [3]

    Robust Fruit Counting: Combining Deep Learning, Tracking, and Structure from Motion,

    X. Liu, S. W. Chen, S. Aditya, N. Sivakumar, S. Dcunha, C. Qu, C. J. Taylor, J. Das and V. Kumar, "Robust Fruit Counting: Combining Deep Learning, Tracking, and Structure from Motion,"arXiv, 2018

    work page 2018

  4. [4]

    Monocular Camera Based Fruit Counting and Mapping With Semantic Data Association,

    X. Liu et al., "Monocular Camera Based Fruit Counting and Mapping With Semantic Data Association,"IEEE Robotics and Automation Letters, vol. 4, pp. 2296-2303, 2019

    work page 2019

  5. [5]

    A comparative study of fruit detection and counting methods for yield mapping in apple orchards,

    N. Häni, P. Roy and V. Isler, "A comparative study of fruit detection and counting methods for yield mapping in apple orchards,"Journal of Field Robotics, vol. 37, pp. 263–282, 2019

    work page 2019

  6. [6]

    FruitNeRF: A Unified Neural Radiance Field based Fruit Counting Framework,

    L. Meyer, A. Gilson, U. Schmidt and M. Stamminger, "FruitNeRF: A Unified Neural Radiance Field based Fruit Counting Framework," arXiv, 2024

    work page 2024

  7. [7]

    PAg-NeRF: Towards fast and efficient end-to-end panoptic 3D representations for agricultural robotics,

    S. Claus, M. Halstead, P. Zimmer, T. Laebe, E. Guclu, C. Stachniss, and C. McCool, "PAg-NeRF: Towards fast and efficient end-to-end panoptic 3D representations for agricultural robotics,"IEEE Robotics and Automation Letters, 2023

    work page 2023

  8. [8]

    Fruit Tracking Over Time Using High-Precision Point Clouds,

    A. Riccardi, S. Kelly, E. Marks, F. Magistri, T. Guadagnino, J. Behley, M. Bennewitz, and C. Stachniss, "Fruit Tracking Over Time Using High-Precision Point Clouds,"Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2023

    work page 2023

Show all 52 references
  1. [9]

    Compact Strawberry Harvesting Tube Employing Laser Cutter,

    M. Sorour, P. J. From, K. Elgeneidy, S. Kanarachos, and M. Sallam, "Compact Strawberry Harvesting Tube Employing Laser Cutter," in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2022

    2022

  2. [10]

    U-Net: Convolutional Net- works for Biomedical Image Segmentation,

    O. Ronneberger, P. Fischer and T. Brox, "U-Net: Convolutional Net- works for Biomedical Image Segmentation,"Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, N. Navab, J. Hornegger, vol. 9351, pp. 234–241, 2015

    2015

  3. [11]

    Complete and accurate holly fruit counting using YOLOX object detection,

    Y. Zhang, W. Zhang, J. Yu, L. He, J. Chen, and Y. He, "Complete and accurate holly fruit counting using YOLOX object detection," Computers and Electronics in Agriculture, vol. 198, p. 107062, 2022

    2022

  4. [12]

    Dimensionality Reduction by Learning an Invariant Mapping,

    R. Hadsell, S. Chopra, and Y. LeCun, "Dimensionality Reduction by Learning an Invariant Mapping,"2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), 2006

    2006

  5. [13]

    A Simple Framework for Contrastive Learning of Visual Representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, "A Simple Framework for Contrastive Learning of Visual Representations," Proceedings of the 37th International Conference on Machine Learning (ICML), 2020

    2020

  6. [14]

    Unsupervised Semantic Segmentation by Distilling Feature Corre- spondences,

    M. Hamilton, Z. Zhang, B. Hariharan, N. Snavely, and W. T. Freeman, "Unsupervised Semantic Segmentation by Distilling Feature Corre- spondences,"International Conference on Learning Representations, 2022

    2022

  7. [15]

    CADet: Fully Self-Supervised Out-Of-Distribution Detection With Contrastive Learning,

    C. Guille-Escuret, P. Rodriguez, D. Vazquez, I. Mitliagkas, and J. Mon- teiro, "CADet: Fully Self-Supervised Out-Of-Distribution Detection With Contrastive Learning,"Advances in Neural Information Processing Systems (NeurIPS), 2024

    2024

  8. [16]

    Blind Localization and Clustering of Anomalies in Textures,

    A.-T. Ardelean and T. Weyrich, "Blind Localization and Clustering of Anomalies in Textures,"Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2024

    2024

  9. [17]

    Contrastive Learning of General-Purpose Audio Representations,

    A. Saeed, D. Grangier, and N. Zeghidour, "Contrastive Learning of General-Purpose Audio Representations,"arXiv preprint, 2020

    2020

  10. [18]

    FaceNet: A Unified Embedding for Face Recognition and Clustering,

    F. Schroff, D. Kalenichenko, and J. Philbin, "FaceNet: A Unified Embedding for Face Recognition and Clustering,"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015

    2015

  11. [19]

    ArcFace: Additive Angular Margin Loss for Deep Face Recognition,

    J. Deng, J. Guo, N. Xue, and S. Zafeiriou, "ArcFace: Additive Angular Margin Loss for Deep Face Recognition,"2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019

    2019

  12. [20]

    Improved Deep Metric Learning with Multi-class N-pair Loss Objective,

    K. Sohn, "Improved Deep Metric Learning with Multi-class N-pair Loss Objective,"Advances in Neural Information Processing Systems (NeurIPS), vol. 29, 2016

    2016

  13. [21]

    Representation Learning with Contrastive Predictive Coding,

    A. van den Oord, Y. Li, and O. Vinyals, "Representation Learning with Contrastive Predictive Coding,"arXiv, 2019

    2019

  14. [22]

    Contrastive Lift: 3D Object Instance Segmentation by Slow-Fast Contrastive Fusion,

    Y. Bhalgat, I. Laina, J. F. Henriques, A. Zisserman and A. Vedaldi, "Contrastive Lift: 3D Object Instance Segmentation by Slow-Fast Contrastive Fusion,"Proceedings of the Thirty-seventh Conference on Neural Information Processing Systems, 2023

    2023

  15. [23]

    NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis,

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoor- thi and R. Ng, "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis,"Commun. ACM, vol. 65, pp. 99–106, Jan. 2022

    2022

  16. [24]

    Momentum contrast for unsupervised visual representation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. "Momentum contrast for unsupervised visual representation learning." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020

    2020

  17. [25]

    Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields,

    J. T. Barron, B. Mildenhall, D. Verbin, P. P. Srinivasan and P. Hed- man, "Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields," Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023

    2023

  18. [26]

    KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs,

    C. Reiser, S. Peng, Y. Liao and A. Geiger, "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs,"CoRR, vol. abs/2103.13744, 2021

    2021 arXiv

  19. [27]

    PlenOctrees for Real-time Rendering of Neural Radiance Fields,

    A. Yu, R. Li, M. Tancik, H. Li, R. Ng and A. Kanazawa, "PlenOctrees for Real-time Rendering of Neural Radiance Fields,"Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021

    2021

  20. [28]

    Plenoxels: Radiance Fields without Neural Networks,

    S. Fridovich-Keil, A. Yu, M. Tancik, Q. Chen, B. Recht and A. Kanazawa, "Plenoxels: Radiance Fields without Neural Networks," Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022

    2022

  21. [29]

    Instant Neural Graphics Primitives with a Multiresolution Hash Encoding,

    T. Müller, A. Evans, C. Schied and A. Keller, "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding,"ACM Transactions on Graphics (TOG), vol. 41, no. 4, pp. 102:1–102:15, July 2022

    2022

  22. [30]

    In-Place Scene Labelling and Understanding with Implicit Scene Representation,

    S. Zhi, T. Laidlow, S. Leutenegger and A. J. Davison, "In-Place Scene Labelling and Understanding with Implicit Scene Representation," CoRR, 2021

    2021

  23. [31]

    Instance Neural Radiance Field,

    Y. Liu, B. Hu, J. Huang, Y.-W. Tai and C.-K. Tang, "Instance Neural Radiance Field,"Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023

    2023

  24. [32]

    Panoptic Lifting for 3D Scene Understanding With Neural Fields,

    Y. Siddiqui, L. Porzi, S. R. Bulò, N. Müller, M. Nießner, A. Dai and P. Kontschieder, "Panoptic Lifting for 3D Scene Understanding With Neural Fields,"Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023

    2023

  25. [33]

    [Online:https:// www.xfrog.com/product-page/library-fruit-trees]

    XFrog Inc., LIBRARY: FRUIT TREES, 2020. [Online:https:// www.xfrog.com/product-page/library-fruit-trees]

    2020

  26. [34]

    BlenderNeRF

    M. Raafat, "BlenderNeRF" (Version 5.0.0), 2023, [Computer software]. https://doi.org/10.5281/zenodo.7926211

    2023 doi

  27. [35]

    Fuji-SfM dataset: A collection of annotated images and point clouds for Fuji apple detection and location using structure-from-motion photogrammetry,

    J. Gené-Mola, R. Sanz-Cortiella, J. R. Rosell-Polo, J. R. Morros, J. Ruiz-Hidalgo, V. Vilaplana and E. Gregorio, "Fuji-SfM dataset: A collection of annotated images and point clouds for Fuji apple detection and location using structure-from-motion photogrammetry,"Data in Brief...

    2020

  28. [36]

    Structure-from-Motion Revisited,

    J. L. Schönberger and J.-M. Frahm, "Structure-from-Motion Revisited," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016

    2016

  29. [37]

    Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks,

    T. Renet al., "Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks,"arXiv, 2024

    2024

  30. [38]

    Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection,

    S. Liuet al., "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection,"arXiv, 2024

    2024

  31. [39]

    Segment Anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y. Lo, P. Dollár and R. Girshick, "Segment Anything,"arXiv, 2023

    2023

  32. [40]

    Segment Anything in High Quality,

    L. Ke, M. Ye, M. Danelljan, Y. Liu, Y.-W. Tai, C.-K. Tang, and F. Yu, "Segment Anything in High Quality,"Proceedings of the Neural Information Processing Systems (NeurIPS), 2023

    2023

  33. [41]

    Detecting Twenty-thousand Classes using Image-level Supervision,

    X. Zhou, R. Girdhar, A. Joulin, P. Krähenbühl and I. Misra, "Detecting Twenty-thousand Classes using Image-level Supervision,"arXiv, 2022

    2022

  34. [42]

    LVIS: A Dataset for Large Vocabulary Instance Segmentation,

    A. Gupta, P. Dollár and R. Girshick, "LVIS: A Dataset for Large Vocabulary Instance Segmentation,"arXiv, 2019

    2019

  35. [43]

    Radford, J

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, G. Krueger, and I. Sutskever, International Conference on Machine Learning (ICML), 2021

    2021

  36. [44]

    Exploring Multi- modal Neural Scene Representations With Applications on Thermal Imaging,

    M. Özer, M. Weiherer, M. Hundhausen, and B. Egger, "Exploring Multi- modal Neural Scene Representations With Applications on Thermal Imaging,"ArXiv, 2024

    2024

  37. [45]

    Nerfstudio: A Modular Framework for Neural Radiance Field Development,

    M. Tancik, E. Weber, E. Ng, R. Li, B. Yi, J. Kerr, T. Wang, A. Kristoffersen, J. Austin, K. Salahi, A. Ahuja, D. McAllister, and A. Kanazawa, "Nerfstudio: A Modular Framework for Neural Radiance Field Development,"ACM SIGGRAPH 2023 Conference, 2023

    2023

  38. [46]

    A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise,

    M. Ester, H.-P. Kriegel, J. Sander and X. Xu, "A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise," inKnowledge Discovery and Data Mining, 1996

    1996

  39. [47]

    Scikit-learn: Machine Learning in Python,

    F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot and É. Duchesnay, "Scikit-learn: Machine Learning in Python,"J. Mach. Learn. Res., vol. ...

    2011

  40. [48]

    Density-Based Clustering Based on Hierarchical Density Estimates,

    R. J. G. B. Campello, D. Moulavi and J. Sander, "Density-Based Clustering Based on Hierarchical Density Estimates," inAdvances in Knowledge Discovery and Data Mining, vol. 7819, pp. 160–172, 2013

    2013

  41. [49]

    Understanding the Behaviour of Contrastive Loss,

    F. Wang and H. Liu, "Understanding the Behaviour of Contrastive Loss,"arXiv, 2021. [Online]. Available: https://arxiv.org/abs/2012.09740

    2021 arXiv

  42. [50]

    B-SLAM-SIM: A Novel Approach to Evaluate the Fusion of Visual SLAM and GPS by Example of Direct Sparse Odometry and Blender,

    A. Kalisz,et al., “B-SLAM-SIM: A Novel Approach to Evaluate the Fusion of Visual SLAM and GPS by Example of Direct Sparse Odometry and Blender,” VISIGRAPP, 2019

    2019

  43. [51]

    3D Gaussian Splatting for Real-Time Radiance Field Rendering,

    B. Kerbl, G. Kopanas, T. Leimkähler, and G. Drettakis, "3D Gaussian Splatting for Real-Time Radiance Field Rendering,"ACM Transactions on Graphics, vol. 42, July 2023

    2023

  44. [52]

    Zero-Shot Object Counting,

    J. Xu, H. Le, V. Nguyen, V. Ranjan, and D. Samaras, n"Zero-Shot Object Counting,"Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 15548–15557, 2023

    2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.