Pith. sign in

REVIEW 5 major objections 6 minor 82 references

FocalClick-XL: Towards Unified and High-quality Interactive Segmentation

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read FocalClick-XL splits interactive segmentation into three pretrained subnets and claims fewer clicks than SAM on every standard benchmark.

desk verdict Strong engineering and useful new benchmarks, but the paper's own Table 7 contradicts the 'state-of-the-art on click-based benchmarks' claim on SBD. read the letter →

arxiv 2506.14686 v1 pith:VUXDBK6D submitted 2025-06-17 cs.CV

Xi Chen , Hengshuang Zhao This is my paper · ORCID
classification cs.CV
keywords interactivesegmentationclick-basedscribbleinteractionboxpromptingcoarsemaskrefinementalphamattingcoarse-to-fineprogressivemagnification
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

FocalClick-XL aims to show that interactive segmentation is better built as three separately trained subnets — one for scene context, one for the target object, and one for fine detail — than as a single monolithic prompt-to-mask model. Each subnet is pretrained on data suited to its job, and only the object-level subnet needs to know what kind of interaction the user gave. If the paper is right, one model can serve clicks, boxes, scribbles, and coarse masks while matching or beating specialized methods on click benchmarks and even predicting alpha mattes for hair and fur. This matters because real annotation tools must handle varied input formats and fine boundaries at low latency, and the paper supplies a concrete architecture plus benchmarks for measuring both.

What carries the argument

The load-bearing mechanism is the three-subnet decomposition. Context-Net reuses a SAM encoder with per-block adapters and runs once per image; Object-Net, a SegFormer-style network, fuses the interaction prompt, encoded as a two-channel positive/negative map, with the zoomed target patch and RoI-aligned context features; Detail-Net, a MobileNetV2 U-shape, takes a tri-map of confident foreground, confident background, and uncertain regions and predicts a refined mask or alpha matte whose per-pixel values encode transparency. Progressive magnification crops small patches around the object and then around low-confidence regions, which keeps the cascaded pipeline fast. Progressive Merge preserves user-annotated details outside the region the new interaction is meant to change.

What would settle it

Run a human annotation study in which real users correct masks on the same images with free-form scribbles and clicks, comparing their interaction counts against the simulator-driven NoS/NoC numbers; a systematic gap beyond the reported noise would refute the transfer claim, as would a failure of the deterministic scribble simulator to predict human placement.

Watch

Extended reading notes

Core claim

The paper claims that decomposing interactive segmentation into context, object, and detail levels, with each subnet pretrained independently, yields state-of-the-art click-based performance and broad interaction support. Concretely, FocalClick-XL reports lower average numbers of clicks to reach 85% and 90% IoU than prior methods, including SAM and SAM-HQ, on GrabCut, Berkeley, SBD, and DAVIS. The same model transfers to box, scribble, and coarse-mask inputs by retraining only a small prompting layer, and the detail subnet trained on matting data outputs alpha values, so prediction is not limited to binary masks. The authors also introduce evaluation protocols and new benchmarks, including DAVIS-585 for correcting flawed initial masks and a deterministic scribble simulator.

Load-bearing premise

The paper's central results rest on the assumption that the simulated users — max-error clicks, Bezier/axial/boundary scribbles, and eroded or dilated coarse masks — behave like real annotators, so performance under simulation transfers to practice.

Editorial extensions

If this is right

  • A new interaction format can be supported by training only its prompting layer, leaving Context-Net and Detail-Net frozen.
  • Because the detail subnet is trained on matting data, the same pipeline outputs alpha mattes for transparent structures, not just binary masks.
  • The DAVIS-585 benchmark turns mask correction into a measurable task, showing large click savings when starting from a flawed initial mask.
  • The deterministic scribble simulator replaces random scribble generation with a reproducible protocol, enabling automatic NoS evaluation.

Reading between the lines

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

  • If real annotators' scribbles and coarse masks resemble the simulated ones, the prompting-layer-only transfer suggests the same decomposition could absorb language or reference-image prompts with a similarly small per-interaction cost; the paper does not test this.
  • The claim that only the object-level subnet is interaction-sensitive implies that interaction types share context and detail knowledge; this could be tested by training on a mixed-interaction dataset and checking whether Object-Net capacity becomes a bottleneck.
  • The DAVIS-585 correction benchmark could be adopted widely, since real annotation workflows often start from prior predictions, though the paper itself only simulates defects on ground-truth masks.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The manuscript extends the FocalClick interactive-segmentation framework into FocalClick-XL, a three-subnet pipeline that decomposes the task into context-level, object-level, and detail-level processing. The Context-Net is built on a SAM encoder with adapters, the Object-Net is a SegFormer-style model that fuses interaction prompts, and the Detail-Net refines low-confidence regions and can predict alpha mattes. The paper reports state-of-the-art click-based results on GrabCut, Berkeley, SBD, and DAVIS, and shows competitive performance for boxes, scribbles, and coarse masks under newly proposed or newly evaluated protocols, including a new DAVIS-585 benchmark for interactive mask correction and a deterministic scribble simulator.

Significance. If the reported results hold, FocalClick-XL is a substantial contribution to interactive segmentation: it combines large-scale pretraining with a coarse-to-fine design, supports multiple interaction formats in one architecture, and introduces reusable evaluation protocols for non-click interactions. The paper's strengths include comparisons against standard external benchmarks, stepwise ablations showing the contribution of each subnet, and the explicit goal of sharing context- and detail-level knowledge across interaction types. The main caveats are that the headline 'state-of-the-art' claim is contradicted by the paper's own SBD numbers, statistical reliability is not established, and the training setup for the XL model is underreported. These issues are fixable and do not invalidate the core methodology.

major comments (5)
  1. [§7.1, Table 7 (SBD block)] The unqualified 'state-of-the-art on click-based benchmarks' claim in the abstract and §7.1 is contradicted by the paper's own Table 7. SimpleClick-H trained on COCO+LVIS achieves SBD NoC85 = 2.85 and NoC90 = 4.70, while FocalClick-XL-H trained on the larger Combined Dataset achieves SBD NoC85 = 3.32 and NoC90 = 5.51. FocalClick-XL is better on GrabCut, Berkeley, and DAVIS, but the SOTA claim should be qualified (e.g., 'on most click-based benchmarks' or 'excluding SBD') and the discrepancy should be discussed rather than glossed over.
  2. [§7.1, Tables 7–8] All quantitative tables report point estimates without error bars, confidence intervals, or significance tests. GrabCut and Berkeley contain only 50 and 100 images, yet several headline comparisons rest on differences of 0.08–0.16 clicks (e.g., FocalClick-XL-B vs FocalClick-XL-H on GrabCut NoC85: 1.30 vs 1.22). Please report per-image variance or paired statistical tests for the central claims, especially in Table 7.
  3. [§7] The training setup for FocalClick-XL is not reported. Whereas §6.1 gives dataset, epoch count, batch size, optimizer, learning-rate schedule, and augmentation for basic FocalClick, §7 does not state the Combined Dataset composition, number of iterations, batch size, learning-rate schedule, input resolutions for the subnets, or the order of click-pretraining versus transfer tuning. Without these details the headline results cannot be reproduced.
  4. [§4.1 and Table 9] The claim that the decomposition 'allows each subnet to undergo scaled pretraining with independent data and supervision' is not directly tested. Table 9 ablates the presence of the three subnets but keeps the pretraining and data scheme fixed; there is no experiment comparing independent per-subnet pretraining with a jointly trained end-to-end counterpart. Please provide such an ablation, or temper the 'maximizing its effectiveness' claim.
  5. [§5.1–§5.2] The scribble, box, and coarse-mask evaluations rest entirely on simulated user interactions, and the only human evidence is a one-sentence user study in §3.2 that is not described in terms of protocol or sample size. Since the paper proposes DAVIS-585 and the scribble protocol as new benchmarks, the simulators should be validated against a small set of real human annotations, or the results should be explicitly framed as simulator-based.
minor comments (6)
  1. [§6.1] The text says 'Adam optimizer of β1 = 0.9, β1 = 0.999'; the second should presumably be β2 = 0.999.
  2. [Table 7] The header lists 'NoC 90' for Berkeley but not 'NoC 85'; please check the column alignment and header formatting so each column is unambiguous.
  3. [§5.1] The text says 'We combine these four strategies' but only three meta-simulators are described (Bezier, axial, boundary scribbles); clarify what the fourth strategy is.
  4. [§4.1 vs Table 14] The text says adapters are added 'after each transformer layer,' while Table 14 uses the label 'Each stage'; specify whether these refer to the same placement and report the number of adapters.
  5. [§5.2] The text promises NoF2085/90 for scribble evaluation, but Table 10 reports only NoS; either add the failure numbers or remove the promise.
  6. [Throughout] There are several typos: 'Comaprisons' in the Table 7 caption, 'Pertub-Level' in Table 12, 'Promtping Layer' in §5.1, and 'archives impressive performance' in §7.1.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FocalClick-XL's claimed gains rest on external benchmarks and independently pretrained subnets, not on self-referential construction.

full rationale

The derivation chain is self-contained. Click-based comparisons in Table 7 use external, independently published benchmarks (GrabCut, Berkeley, SBD, DAVIS); training uses external datasets (COCO, LVIS, SBD, synthetic/matting data). The self-citation to FocalClick [9] is an architectural lineage, but the paper re-describes the pipeline and reports fresh experiments rather than invoking an unverified prior theorem. The newly proposed DAVIS-585, scribble protocol, and coarse-mask perturbation sets are author-defined evaluation instruments; they are not optimized against in the training loss, so they do not make the reported numbers true by construction. No fitted parameter is renamed as a prediction, and no uniqueness result is imported from the authors' earlier work. The apparent tension between the abstract's state-of-the-art claim and the SBD NoC85 value in Table 7 (SimpleClick-H 2.85 vs FocalClick-XL-H 3.32) is an internal-consistency/correctness issue, not a circularity, because the evaluation still measures the method on external data. Under the stated rules, no circular step can be quoted, so the appropriate score is 0.

Assumptions & free parameters 6 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new physical or postulated entities. Its new artifacts are benchmarks, simulators, and a network architecture, which are not entities in the ledger sense. The central empirical claims rest mainly on the transferability of SAM features, the realism of simulated interactions, and the usefulness of matting supervision for boundary details.

free parameters (6)
  • Target Crop expand ratio r_TC = 1.4
    Expansion factor for the crop around the target object; copied from prior works and hand-set, affects how much context the Object-Net sees.
  • Focus Crop expand ratio r_FC = 1.4
    Expansion factor for the detail patch; hand-set, affects the Detail-Net input region.
  • Defect type probabilities for DAVIS-585 = [0.65, 0.25, 0.1]
    Probabilities for boundary error, external false positive, and internal true negative in the simulated flawed masks; chosen by the authors after informal observation of real tasks.
  • Initial mask IOU range for DAVIS-585 = [0.75, 0.85]
    Range used to generate the initial masks; derived from a small user study where users discarded masks below 75% IOU.
  • Input resolutions for subnets = Context-Net 1024x1024, Object-Net 384x384, Detail-Net 256x256
    Hand-set trade-off between speed and quality in the progressive magnification strategy.
  • Binarization threshold for alpha matte = 0.5
    Threshold used to convert Detail-Net transparency predictions into binary masks; standard but directly affects reported IoU values.
assumptions (3)
  • domain assumption SAM's pretrained image encoder, with lightweight adapters, provides transferable context and object priors for interactive segmentation.
    Context-Net initializes from the SAM encoder and only adapter layers are optimized; if SAM features do not transfer to interactive segmentation, the reported gains would disappear. Ablations in Tab. 14 support the choice but do not prove transfer across arbitrary domains.
  • domain assumption Simulated user interactions (max-error clicks, meta-simulator scribbles, perturbed coarse masks) are representative of real annotator behavior.
    All training and most evaluation use synthetic interactions generated by the algorithms in Sec. 5. The authors observe real error distributions informally for DAVIS-585, but there is no systematic validation against human annotation traces.
  • domain assumption Matting supervision on alpha values transfers to boundary refinement for binary segmentation.
    Detail-Net is trained with MSE loss on transparency values from matting datasets; the paper assumes this improves binary mask quality, but it reports no direct matting evaluation to confirm the alpha-matte capability.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FocalClick-XL: Towards Unified and High-quality Interactive Segmentation." pith.science (2026). https://pith.science/paper/VUXDBK6D

@misc{pith2026250614686,
  author       = {Pith},
  title        = {Pith review of: FocalClick-XL: Towards Unified and High-quality Interactive Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VUXDBK6D}},
  note         = {Machine review of arXiv:2506.14686}
}
read the original abstract

Interactive segmentation enables users to extract binary masks of target objects through simple interactions such as clicks, scribbles, and boxes. However, existing methods often support only limited interaction forms and struggle to capture fine details. In this paper, we revisit the classical coarse-to-fine design of FocalClick and introduce significant extensions. Inspired by its multi-stage strategy, we propose a novel pipeline, FocalClick-XL, to address these challenges simultaneously. Following the emerging trend of large-scale pretraining, we decompose interactive segmentation into meta-tasks that capture different levels of information -- context, object, and detail -- assigning a dedicated subnet to each level.This decomposition allows each subnet to undergo scaled pretraining with independent data and supervision, maximizing its effectiveness. To enhance flexibility, we share context- and detail-level information across different interaction forms as common knowledge while introducing a prompting layer at the object level to encode specific interaction types. As a result, FocalClick-XL achieves state-of-the-art performance on click-based benchmarks and demonstrates remarkable adaptability to diverse interaction formats, including boxes, scribbles, and coarse masks. Beyond binary mask generation, it is also capable of predicting alpha mattes with fine-grained details, making it a versatile and powerful tool for interactive segmentation.

Figures

Figures reproduced from arXiv: 2506.14686 by the authors.

Figure 1
Figure 1. Demonstrations for FocalClick-XL . Our method is compatible with various formats of user interactions like clicks, scribbles, boxes, coarse masks, etc., and could predict highly￾refined details for both transparent and solid objects. features for each part of the image and leaves a small prompt encoder to model each interaction. Besides, SAM only supports click and box, and it is hard for the current prompt encoder … view at source ↗
Figure 2
Figure 2. Overall framework of FocalClick. We take the image, two click maps, and the previous mask as input. We use binary disks with radius 2 to represent the click. First, we select the Target Crop around the target object and resize it to a small size. It is then sent into Segmentor to predict a coarse mask. Next, we chose a Focus Crop by calculating the different regions between the previous masks and the coarse predicti… view at source ↗
Figure 3
Figure 3. The extended framework of FocalClick-XL . [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Simulated user interactions, such as scribbles, boxes, and coarse masks. For the scribbles, we first develop several meta-simulators and compose them for more versatile results. tasks. The Prompting Layer of the Object-Net projects the interaction maps into the control…
Figure 5
Figure 5. Figure 5: The procedure of deterministic scribble generation. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Qualitative analysis for the effectiveness of Refiner. The first column denotes the Target Crop in yellow and the Focus Crop in red. The second and the third column demonstrate the mask in focus crop before and after refinement. TABLE 6: Computation analysis for FocalC…
Figure 7
Figure 7. Figure 7: Demonstrations for FocalClick-XL . Our method provides a unified solution for various interaction formats and predicts high-quality masks with fine details. IoU: 84.5 IoU: 84.1 IoU: 60.5 IoU: 92.0 IoU: 93.9 IoU: 87.4 SAM DTC IoU: 96.1 IoU: 80.9 [PITH_FULL_IMAGE:figure…
Figure 8
Figure 8. Figure 8: Qualitative comparisons with SAM. FocalClick￾XL shows significantly better mask qualities compared with SAM given a single click. 8 CONCLUSION In this paper, we build upon the classical interactive segmentation framework, FocalClick, and introduce sig￾nificant extensio…
Figure 9
Figure 9. Figure 9: More visualization results for FocalClick-XL . [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

82 extracted references · 60 canonical work pages

  1. [1]

    Interactive object segmentation with inside-outside guidance

    Shiyin Zhang, Jun Hao Liew, Yunchao Wei, Shikui Wei, and Yao Zhao. Interactive object segmentation with inside-outside guidance. InCVPR, 2020. 1, 2 13 BoxCoarseMask Click IoU:95.3 IoU:97.2 IoU:83.1 IoU:90.1 IoU:95.7 IoU:96.6 IoU:97.6 IoU:94.6 IoU:88.7 IoU:96.1 IoU:97.7 IoU:86.7 IoU:96.1 IoU:96.9 IoU:96.1 IoU:89.7 IoU:59.2 IoU:29.0 Scribble IoU:90.9 IoU:88...

    2020

  2. [2]

    Segment anything

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alexander C Berg, Wan-Yen Lo, et al. Segment anything. arXiv:2304.02643, 2023. 1, 2, 6, 9, 10, 11

    arXiv 2023

  3. [3]

    Error-tolerant scribbles based interactive image segmentation

    Junjie Bai and Xiaodong Wu. Error-tolerant scribbles based interactive image segmentation. InCVPR, 2014. 1, 2

    2014

  4. [4]

    Efficient Full Image Interactive Segmentation by Leveraging Within-image Appearance Similarity

    Mykhaylo Andriluka, Stefano Pellegrini, Stefan Popov, and Vit- torio Ferrari. Efficient full image interactive segmentation by 14 leveraging within-image appearance similarity.arXiv:2007.08173,

    work page Pith review arXiv 2007

  5. [5]

    Deepigeos: a deep interactive geodesic framework for medical image segmentation.TP AMI,

    Guotai Wang, Maria A Zuluaga, Wenqi Li, Rosalind Pratt, Pre- mal A Patel, Michael Aertsen, Tom Doel, Anna L David, Jan Deprest, S ´ebastien Ourselin, et al. Deepigeos: a deep interactive geodesic framework for medical image segmentation.TP AMI,

    work page

  6. [6]

    Eirikur Agustsson, Jasper R. R. Uijlings, and Vittorio Ferrari. Interactive full image segmentation by considering all regions jointly. InCVPR, 2019. 1, 2, 6, 9

    work page 2019

  7. [7]

    Reviving iterative training with mask guidance for interactive segmenta- tion.arXiv:2102.06583, 2021

    Konstantin Sofiiuk, Ilia A Petrov, and Anton Konushin. Reviving iterative training with mask guidance for interactive segmenta- tion.arXiv:2102.06583, 2021. 1, 2, 3, 4, 6, 7, 8, 9, 10

    arXiv 2021

  8. [8]

    Simpleclick: Interactive image segmentation with simple vision transformers

    Qin Liu, Zhenlin Xu, Gedas Bertasius, and Marc Niethammer. Simpleclick: Interactive image segmentation with simple vision transformers. InICCV, 2023. 1, 2, 6, 10, 11

    work page 2023

Show all 82 references
  1. [9]

    Focalclick: Towards practical interactive image segmentation

    Xi Chen, Zhiyan Zhao, Yilei Zhang, Manni Duan, Donglian Qi, and Hengshuang Zhao. Focalclick: Towards practical interactive image segmentation. InCVPR, 2022. 1, 2, 6

    2022

  2. [10]

    Conditional diffusion for interactive segmentation

    Xi Chen, Zhiyan Zhao, Feiwu Yu, Yilei Zhang, and Manni Duan. Conditional diffusion for interactive segmentation. InICCV, 2021. 1, 2, 8, 9, 10, 11

    2021

  3. [11]

    Iteratively trained interactive segmentation

    Sabarinath Mahadevan, Paul Voigtlaender, and Bastian Leibe. Iteratively trained interactive segmentation. InBMVC, 2018. 1, 6, 7

    2018

  4. [12]

    Interactive image segmentation with first click attention

    Zheng Lin, Zhao Zhang, Lin-Zhuo Chen, Ming-Ming Cheng, and Shao-Ping Lu. Interactive image segmentation with first click attention. InCVPR, 2020. 1, 2, 8, 9, 10

    2020

  5. [13]

    Edge- flow: Achieving practical interactive segmentation with edge- guided flow

    Yuying Hao, Yi Liu, Zewu Wu, Lin Han, Yizhou Chen, Guowei Chen, Lutao Chu, Shiyu Tang, Zhiliang Yu, Zeyu Chen, et al. Edge- flow: Achieving practical interactive segmentation with edge- guided flow. InICCV, 2021. 1, 2, 9, 10

    2021

  6. [14]

    f-brs: Rethinking backpropagating refinement for interactive segmentation

    Konstantin Sofiiuk, Ilia Petrov, Olga Barinova, and Anton Konushin. f-brs: Rethinking backpropagating refinement for interactive segmentation. InCVPR, 2020. 1, 2, 3, 8, 9, 10, 11

    2020

  7. [15]

    ” grabcut” interactive foreground extraction using iterated graph cuts.TOG, 2004

    Carsten Rother, Vladimir Kolmogorov, and Andrew Blake. ” grabcut” interactive foreground extraction using iterated graph cuts.TOG, 2004. 2, 4, 9, 10

    2004

  8. [16]

    Geodesic star convexity for interactive image segmentation

    Varun Gulshan, Carsten Rother, Antonio Criminisi, Andrew Blake, and Andrew Zisserman. Geodesic star convexity for interactive image segmentation. InCVPR, 2010. 2, 10

    2010

  9. [17]

    Random walks for image segmentation.TP AMI, 2006

    Leo Grady. Random walks for image segmentation.TP AMI, 2006. 2, 10

    2006

  10. [18]

    Nonparametric higher-order learning for interactive segmentation

    Tae Hoon Kim, Kyoung Mu Lee, and Sang Uk Lee. Nonparametric higher-order learning for interactive segmentation. InCVPR, 2010. 2

    2010

  11. [19]

    Deep interactive object selection

    Ning Xu, Brian Price, Scott Cohen, Jimei Yang, and Thomas S Huang. Deep interactive object selection. InCVPR, 2016. 2, 7, 8, 10

    2016

  12. [20]

    Interactive image segmentation with latent diversity

    Zhuwen Li, Qifeng Chen, and Vladlen Koltun. Interactive image segmentation with latent diversity. InCVPR, 2018. 2, 4, 8, 10

    2018

  13. [21]

    Multiseg: Semantically meaningful, scale-diverse segmentations from minimal user input

    Jun Hao Liew, Scott Cohen, Brian Price, Long Mai, Sim-Heng Ong, and Jiashi Feng. Multiseg: Semantically meaningful, scale-diverse segmentations from minimal user input. InICCV, 2019. 2

    2019

  14. [22]

    Interactive image segmenta- tion via backpropagating refinement scheme

    Won-Dong Jang and Chang-Su Kim. Interactive image segmenta- tion via backpropagating refinement scheme. InCVPR, 2019. 2, 8, 10

    2019

  15. [23]

    Scrib- blesup: Scribble-supervised convolutional networks for semantic segmentation

    Di Lin, Jifeng Dai, Jiaya Jia, Kaiming He, and Jian Sun. Scrib- blesup: Scribble-supervised convolutional networks for semantic segmentation. InCVPR, 2016. 2

    2016

  16. [24]

    Robust interactive image segmen- tation via graph-based manifold ranking

    Enhua Wu Hong Li, Wen Wu. Robust interactive image segmen- tation via graph-based manifold ranking. InComputational Visual Media, 2015. 2

    2015

  17. [25]

    Scribble based interactive page layout segmentation using gabor filter

    Majeed Kassis and Jihad El-Sana. Scribble based interactive page layout segmentation using gabor filter. InICFHR, 2016. 2

    2016

  18. [26]

    Semantic-sam: Segment and recognize anything at any granularity

    Feng Li, Hao Zhang, Peize Sun, Xueyan Zou, Shilong Liu, Jianwei Yang, Chunyuan Li, Lei Zhang, and Jianfeng Gao. Semantic-sam: Segment and recognize anything at any granularity. arXiv:2307.04767, 2023. 2

    2023 arXiv

  19. [27]

    Samus: Adapting segment anything model for clinically-friendly and generalizable ultrasound image segmenta- tion.arXiv:2309.06824, 2023

    Xian Lin, Yangyang Xiang, Li Zhang, Xin Yang, Zengqiang Yan, and Li Yu. Samus: Adapting segment anything model for clinically-friendly and generalizable ultrasound image segmenta- tion.arXiv:2309.06824, 2023. 2

    2023 arXiv

  20. [28]

    Segment and track anything

    Yangming Cheng, Liulei Li, Yuanyou Xu, Xiaodi Li, Zongxin Yang, Wenguan Wang, and Yi Yang. Segment and track anything. arXiv:2305.06558, 2023. 2

    2023 arXiv

  21. [29]

    Track anything: Segment anything meets videos

    Jinyu Yang, Mingqi Gao, Zhe Li, Shang Gao, Fangjing Wang, and Feng Zheng. Track anything: Segment anything meets videos. arXiv:2304.11968, 2023. 2

    2023 arXiv

  22. [30]

    Sam3d: Segment anything in 3d scenes.arXiv:2306.03908,

    Yunhan Yang, Xiaoyang Wu, Tong He, Hengshuang Zhao, and Xi- hui Liu. Sam3d: Segment anything in 3d scenes.arXiv:2306.03908,

    arXiv

  23. [31]

    Sam3d: Zero-shot 3d object detection via segment anything model.arXiv:2306.02245,

    Dingyuan Zhang, Dingkang Liang, Hongcheng Yang, Zhikang Zou, Xiaoqing Ye, Zhe Liu, and Xiang Bai. Sam3d: Zero-shot 3d object detection via segment anything model.arXiv:2306.02245,

    arXiv

  24. [32]

    Caption anything: Interactive image description with diverse multimodal controls.arXiv:2305.02677, 2023

    Teng Wang, Jinrui Zhang, Junjie Fei, Yixiao Ge, Hao Zheng, Yunlong Tang, Zhe Li, Mingqi Gao, Shanshan Zhao, Ying Shan, et al. Caption anything: Interactive image description with diverse multimodal controls.arXiv:2305.02677, 2023. 2

    2023 arXiv

  25. [33]

    Inpaint anything: Segment anything meets image inpainting.arXiv:2304.06790, 2023

    Tao Yu, Runseng Feng, Ruoyu Feng, Jinming Liu, Xin Jin, Wenjun Zeng, and Zhibo Chen. Inpaint anything: Segment anything meets image inpainting.arXiv:2304.06790, 2023. 2

    2023 arXiv

  26. [34]

    Segment anything in high quality

    Lei Ke, Mingqiao Ye, Martin Danelljan, Yifan Liu, Yu-Wing Tai, Chi-Keung Tang, and Fisher Yu. Segment anything in high quality. arXiv:2306.01567, 2023. 2, 9, 10, 11

    2023 arXiv

  27. [35]

    Segment everything everywhere all at once.arXiv:2304.06718, 2023

    Xueyan Zou, Jianwei Yang, Hao Zhang, Feng Li, Linjie Li, Jianfeng Gao, and Yong Jae Lee. Segment everything everywhere all at once.arXiv:2304.06718, 2023. 2

    2023 arXiv

  28. [36]

    Sam-adapter: Adapting segment anything in underperformed scenes

    Tianrun Chen, Lanyun Zhu, Chaotao Deng, Runlong Cao, Yan Wang, Shangzhan Zhang, Zejian Li, Lingyun Sun, Ying Zang, and Papa Mao. Sam-adapter: Adapting segment anything in underperformed scenes. InICCV, 2023. 2

    2023

  29. [37]

    Sam fails to segment anything?–sam-adapter: Adapting sam in underperformed scenes: Camouflage, shadow, and more.arXiv:2304.09148, 2023

    Tianrun Chen, Lanyun Zhu, Chaotao Ding, Runlong Cao, Shangzhan Zhang, Yan Wang, Zejian Li, Lingyun Sun, Papa Mao, and Ying Zang. Sam fails to segment anything?–sam-adapter: Adapting sam in underperformed scenes: Camouflage, shadow, and more.arXiv:2304.09148, 2023. 2

    2023 arXiv

  30. [38]

    Medical sam adapter: Adapting segment anything model for medical image segmenta- tion.arXiv:2304.12620, 2023

    Junde Wu, Rao Fu, Huihui Fang, Yuanpei Liu, Zhaowei Wang, Yanwu Xu, Yueming Jin, and Tal Arbel. Medical sam adapter: Adapting segment anything model for medical image segmenta- tion.arXiv:2304.12620, 2023. 2

    2023 arXiv

  31. [39]

    How segment anything model (sam) boost medical image segmentation?arXiv:2305.03678, 2023

    Yichi Zhang and Rushi Jiao. How segment anything model (sam) boost medical image segmentation?arXiv:2305.03678, 2023. 2

    2023 arXiv

  32. [40]

    Ma-sam: Modality-agnostic sam adaptation for 3d medical image segmentation.arXiv:2309.08842, 2023

    Cheng Chen, Juzheng Miao, Dufan Wu, Zhiling Yan, Sekeun Kim, Jiang Hu, Aoxiao Zhong, Zhengliang Liu, Lichao Sun, Xiang Li, et al. Ma-sam: Modality-agnostic sam adaptation for 3d medical image segmentation.arXiv:2309.08842, 2023. 2

    2023 arXiv

  33. [41]

    Regional interactive image segmentation networks

    JunHao Liew, Yunchao Wei, Wei Xiong, Sim-Heng Ong, and Jiashi Feng. Regional interactive image segmentation networks. InICCV,

  34. [42]

    Getting to 99% accuracy in interactive segmentation

    Marco Forte, Brian Price, Scott Cohen, Ning Xu, and Franc ¸ois Piti´e. Getting to 99% accuracy in interactive segmentation. arXiv:2003.07932, 2020. 2, 9, 10

    2003 arXiv

  35. [43]

    Pyramid scene parsing network

    Hengshuang Zhao, Jianping Shi, Xiaojuan Qi, Xiaogang Wang, and Jiaya Jia. Pyramid scene parsing network. InCVPR, 2017. 3

    2017

  36. [44]

    Deeplab: Semantic image segmenta- tion with deep convolutional nets, atrous convolution, and fully connected crfs.TP AMI, 2017

    Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, and Alan L Yuille. Deeplab: Semantic image segmenta- tion with deep convolutional nets, atrous convolution, and fully connected crfs.TP AMI, 2017. 3

    2017

  37. [45]

    Fully convolutional networks for semantic segmentation

    Jonathan Long, Evan Shelhamer, and Trevor Darrell. Fully convolutional networks for semantic segmentation. InCVPR,

  38. [46]

    Unified perceptual parsing for scene understanding

    Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, and Jian Sun. Unified perceptual parsing for scene understanding. InECCV,

  39. [47]

    Large kernel matters–improve semantic segmentation by global convolutional network

    Chao Peng, Xiangyu Zhang, Gang Yu, Guiming Luo, and Jian Sun. Large kernel matters–improve semantic segmentation by global convolutional network. InCVPR, 2017. 3

    2017

  40. [48]

    Deep high-resolution representation learning for visual recognition.TP AMI, 2020

    Jingdong Wang, Ke Sun, Tianheng Cheng, Borui Jiang, Chaorui Deng, Yang Zhao, Dong Liu, Yadong Mu, Mingkui Tan, Xinggang Wang, et al. Deep high-resolution representation learning for visual recognition.TP AMI, 2020. 3, 7

    2020

  41. [49]

    Object-contextual representations for semantic segmentation

    Yuhui Yuan, Xilin Chen, and Jingdong Wang. Object-contextual representations for semantic segmentation. InECCV, 2020. 3

    2020

  42. [50]

    Segformer: Simple and efficient design for semantic segmentation with transformers

    Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M Alvarez, and Ping Luo. Segformer: Simple and efficient design for semantic segmentation with transformers. InNeurIPS, 2021. 3, 5, 7

    2021

  43. [51]

    Mask r-cnn

    Kaiming He, Georgia Gkioxari, Piotr Doll ´ar, and Ross Girshick. Mask r-cnn. InICCV, 2017. 3 15

    2017

  44. [52]

    Xception: Deep learning with depthwise separa- ble convolutions

    Franc ¸ois Chollet. Xception: Deep learning with depthwise separa- ble convolutions. InCVPR, 2017. 3

    2017

  45. [53]

    A comparative evalu- ation of interactive segmentation algorithms.Pattern Recognition,

    Kevin McGuinness and Noel E O’connor. A comparative evalu- ation of interactive segmentation algorithms.Pattern Recognition,

  46. [54]

    A benchmark dataset and evaluation methodology for video object segmentation

    Federico Perazzi, Jordi Pont-Tuset, Brian McWilliams, Luc Van Gool, Markus Gross, and Alexander Sorkine-Hornung. A benchmark dataset and evaluation methodology for video object segmentation. InCVPR, 2016. 4, 9, 10

    2016

  47. [55]

    Semantic contours from inverse detectors

    Bharath Hariharan, Pablo Arbel ´aez, Lubomir Bourdev, Subhransu Maji, and Jitendra Malik. Semantic contours from inverse detectors. InICCV. IEEE, 2011. 4, 7, 9, 10

    2011

  48. [56]

    Mask r-cnn

    Kaiming He, Georgia Gkioxari, Piotr Doll ´ar, and Ross Girshick. Mask r-cnn. InICCV, 2017. 5

    2017

  49. [57]

    Context-aware image matting for simultaneous foreground and alpha estimation

    Qiqi Hou and Feng Liu. Context-aware image matting for simultaneous foreground and alpha estimation. InICCV, 2019. 5

    2019

  50. [58]

    Deep image matting

    Ning Xu, Brian Price, Scott Cohen, and Thomas Huang. Deep image matting. InCVPR, 2017. 5

    2017

  51. [59]

    Mobilenetv2: Inverted residuals and linear bottlenecks

    Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmogi- nov, and Liang-Chieh Chen. Mobilenetv2: Inverted residuals and linear bottlenecks. InCVPR, 2018. 5

    2018

  52. [60]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll ´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. InECCV, 2014. 5, 7, 9, 10

    2014

  53. [61]

    Lvis: A dataset for large vocabulary instance segmentation

    Agrim Gupta, Piotr Dollar, and Ross Girshick. Lvis: A dataset for large vocabulary instance segmentation. InCVPR, 2019. 5, 7, 9, 10

    2019

  54. [62]

    Salient object detection: A benchmark.TIP, 2015

    Ali Borji, Ming-Ming Cheng, Huaizu Jiang, and Jia Li. Salient object detection: A benchmark.TIP, 2015. 5, 7, 10

    2015

  55. [63]

    Scene parsing through ade20k dataset

    Bolei Zhou, Hang Zhao, Xavier Puig, Sanja Fidler, Adela Barriuso, and Antonio Torralba. Scene parsing through ade20k dataset. In CVPR, 2017. 5, 7, 10

    2017

  56. [64]

    Learning to detect salient objects with image-level supervision

    Lijun Wang, Huchuan Lu, Yifan Wang, Mengyang Feng, Dong Wang, Baocai Yin, and Xiang Ruan. Learning to detect salient objects with image-level supervision. InCVPR, 2017. 5, 7, 10

    2017

  57. [65]

    Youtube-vos: A large-scale video object segmentation benchmark.arXiv preprint arXiv:1809.03327, 2018

    Ning Xu, Linjie Yang, Yuchen Fan, Dingcheng Yue, Yuchen Liang, Jianchao Yang, and Thomas Huang. Youtube-vos: A large-scale video object segmentation benchmark.arXiv preprint arXiv:1809.03327, 2018. 5, 7, 10

    2018 arXiv

  58. [66]

    Deep interactive thin object selection

    Jun Hao Liew, Scott Cohen, Brian Price, Long Mai, and Jiashi Feng. Deep interactive thin object selection. InWACV, 2021. 5, 7, 10

    2021

  59. [67]

    Dovenet: Deep image harmoniza- tion via domain verification

    Wenyan Cong, Jianfu Zhang, Li Niu, Liu Liu, Zhixin Ling, Weiyuan Li, and Liqing Zhang. Dovenet: Deep image harmoniza- tion via domain verification. InCVPR, 2020. 5, 7, 10

    2020

  60. [68]

    Bridging composite and real: towards end-to-end deep image matting.IJCV, 2022

    Jizhizi Li, Jing Zhang, Stephen J Maybank, and Dacheng Tao. Bridging composite and real: towards end-to-end deep image matting.IJCV, 2022. 5

    2022

  61. [69]

    Attention-guided hierarchical structure aggregation for image matting

    Yu Qiao, Yuhao Liu, Xin Yang, Dongsheng Zhou, Mingliang Xu, Qiang Zhang, and Xiaopeng Wei. Attention-guided hierarchical structure aggregation for image matting. InCVPR, 2020. 5

    2020

  62. [70]

    Mobilenets: Efficient convolutional neural networks for mobile vision applications.arXiv:1704.04861, 2017

    Andrew G Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, and Hartwig Adam. Mobilenets: Efficient convolutional neural networks for mobile vision applications.arXiv:1704.04861, 2017. 8

    2017 arXiv

  63. [71]

    Efficientnet: Rethinking model scaling for convolutional neural networks

    Mingxing Tan and Quoc Le. Efficientnet: Rethinking model scaling for convolutional neural networks. InICML, 2019. 8

    2019

  64. [72]

    Deep image matting

    Ning Xu, Brian Price, Scott Cohen, and Thomas Huang. Deep image matting. InCVPR, 2017. 10

    2017

  65. [73]

    The synthesizability of texture examples

    Dengxin Dai, Hayko Riemenschneider, and Luc Van Gool. The synthesizability of texture examples. InCVPR, 2014. 10

    2014

  66. [74]

    Interactive graph cuts for optimal boundary & region segmentation of objects in nd images

    Yuri Y Boykov and M-P Jolly. Interactive graph cuts for optimal boundary & region segmentation of objects in nd images. InICCV,

  67. [75]

    Content-aware multi-level guidance for interactive instance segmentation

    Soumajit Majumder and Angela Yao. Content-aware multi-level guidance for interactive instance segmentation. InCVPR, 2019. 10

    2019

  68. [76]

    The pascal visual object classes (voc) challenge.IJCV, (2), 2010

    Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. The pascal visual object classes (voc) challenge.IJCV, (2), 2010. 10

    2010

  69. [77]

    Pseu- doclick: Interactive image segmentation with click imitation

    Qin Liu, Meng Zheng, Benjamin Planche, Srikrishna Karanam, Terrence Chen, Marc Niethammer, and Ziyan Wu. Pseu- doclick: Interactive image segmentation with click imitation. arXiv:2207.05282, 2022. 10

    2022 arXiv

  70. [78]

    Interformer: Real-time interactive image segmentation

    You Huang, Hao Yang, Ke Sun, Shengchuan Zhang, Liujuan Cao, Guannan Jiang, and Rongrong Ji. Interformer: Real-time interactive image segmentation. InICCV, 2023. 10

    2023

  71. [79]

    Segmented anisotropic sstem dataset of neural tissue.figshare, 2013

    Stephan Gerhard, Jan Funke, Julien Martel, Albert Cardona, and Richard Fetter. Segmented anisotropic sstem dataset of neural tissue.figshare, 2013. 9, 10

    2013

  72. [80]

    The rsna- asnr-miccai brats 2021 benchmark on brain tumor segmentation and radiogenomic classification.arXiv:2107.02314, 2021

    Ujjwal Baid, Satyam Ghodasara, Suyash Mohan, Michel Bilello, Evan Calabrese, Errol Colak, Keyvan Farahani, Jayashree Kalpathy-Cramer, Felipe C Kitamura, Sarthak Pati, et al. The rsna- asnr-miccai brats 2021 benchmark on brain tumor segmentation and radiogenomic classification....

    2021 arXiv

  73. [81]

    Cascadepsp: Toward class-agnostic and very high-resolution segmentation via global and local refinement

    Ho Kei Cheng, Jihoon Chung, Yu-Wing Tai, and Chi-Keung Tang. Cascadepsp: Toward class-agnostic and very high-resolution segmentation via global and local refinement. InCVPR, 2020. 10, 11

    2020

  74. [82]

    Unidenti- fied video objects: A benchmark for dense, open-world segmenta- tion

    Weiyao Wang, Matt Feiszli, Heng Wang, and Du Tran. Unidenti- fied video objects: A benchmark for dense, open-world segmenta- tion. InICCV, 2021. 10, 11

    2021

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.