Pith. sign in

REVIEW 2 major objections 4 minor 33 references

Open-Set Object Detection By Aligning Known Class Representations

T0 review · 2 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that aligning known-class proposals to semantic text embeddings, decorrelating clusters, and learning objectness from centerness lets an open-set detector correctly mark unseen objects as unknown instead of…

desk verdict Sensible OSOD combination with a decent ablation story, but the headline gains are uninterpretable because the entropy-threshold post-processing is applied to the proposed model and apparently not to the baselines. read the letter →

arxiv 2412.20701 v1 pith:RNJHSSTP submitted 2024-12-30 cs.CV

Hiran Sarkar , Vishal Chudasama , Naoyuki Onoe , Pankaj Wasnik , Vineeth N Balasubramanian This is my paper · ORCID
classification cs.CV
keywords open-setobjectdetectionsemanticclusteringclassdecorrelationfocuslossentropythresholdingharmonicmeanprecision
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tackles open-set object detection: finding objects from classes never seen during training and marking them 'unknown' rather than forcing them into a known category. Its central claim is that prior contrastive-clustering detectors fail on unknowns that are semantically close to a known class (a zebra called a horse) because their clusters are not aligned to semantic meaning. The proposed fix aligns region-proposal features with CLIP text embeddings of class names, enforces orthogonality between class clusters, and learns objectness from centerness and geometry rather than from class labels alone. The authors report large gains in unknown-detection precision (APu) and lower misclassification (AOSE, WI) on MS-COCO and PASCAL VOC, and introduce a combined metric, harmonic mean precision, to summarize known-plus-unknown performance.

What carries the argument

The load-bearing mechanism is the semantic clustering module: a cross-entropy loss that pulls each ROI-aligned proposal feature toward the CLIP text embedding of its ground-truth class name, so clusters form around semantic axes (Eq. 1–2). Around it sit two supporting modules: a class decorrelation loss that samples one feature per class per batch and diagonalizes their cosine-similarity matrix via a softmax cross-entropy term (Eq. 3–5), and an object focus loss that replaces pure classification-based objectness with the geometric mean of a centerness loss and the standard objectness loss (Eq. 6–7), so the RPN learns generic 'thingness' rather than memorizing training categories. At evaluation, an entropy threshold (0.85) relabels low-confidence known-class predictions as 'unknown', and a new metric, HMP, is defined as the harmonic mean of known mAP and unknown APu.

What would settle it

Run the OpenDet baseline with the same entropy-thresholding post-processing at 0.85 on VOC-COCO-40; if its APu rises to near the proposed model's 14.00, the reported gains are mostly an artifact of the evaluation rule rather than the learned clusters. Alternatively, evaluate the proposed model with entropy thresholding disabled and compare APu to OpenDet without it.

Watch

Extended reading notes

Core claim

In the authors' telling, the reason open-set detectors overconfidently mislabel unknowns is that their known-class feature clusters are arranged by instance-level contrastive learning, which does not respect semantic class boundaries. The paper's discovery is that replacing that implicit clustering with an explicit semantic alignment—cosine-similarity cross-entropy between proposal features and CLIP text embeddings of the class names—together with a softmax-form orthogonality constraint on sampled class features and a geometric-mean object focus loss, yields well-separated clusters that leave room for unknown objects to be detected as unknowns. On the VOC-COCO-T1 and T2 protocols with ResNet50, ConvNet, and Swin-T backbones, the proposed model reports APu improvements of roughly 4–18 points over the OpenDet baseline and consistent reductions in Wilderness Impact and Absolute Open-Set Error, with a modest known-class mAP trade-off.

Load-bearing premise

The evaluations assume that applying the entropy-thresholding relabeling (threshold 0.85, picked on the VOC-COCO-40 test set) only to the proposed model—while baselines are evaluated without it—does not inflate the reported improvements in APu, WI, and AOSE.

Editorial extensions

If this is right

  • If the alignment claim holds, open-set detectors trained on the same closed-set data can be made markedly safer for robotics and autonomous driving, where mislabeling a novel obstacle as a known class is more dangerous than saying 'unknown'.
  • The entropy-thresholding evaluation rule, applied post-hoc, turns any detector into a better open-set detector; the paper's reported gains should be understood as combining model improvements with this evaluation change.
  • The HMP metric gives the community a single number that punishes detectors which ignore unknowns (APu=0 gives HMP=0), so future OSOD papers can compare trade-offs on one axis.
  • Semantic alignment via frozen text embeddings may transfer to other open-set tasks like incremental detection and open-set domain adaptation, as the authors note.
  • The known-class mAP cost is small but nonzero in several settings, suggesting the approach trades a little closed-set accuracy for open-set safety.

Reading between the lines

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

  • Because the threshold 0.85 was selected on the VOC-COCO-40 test set (Supplementary B.2), the absolute APu numbers are optimistic; a fair cross-method comparison should apply the same entropy rule to baselines or tune thresholds on a held-out split. This is our editorial caution, not the paper's claim.
  • The decorrelation module samples one feature per class per batch; in large-batch settings this may underrepresent intra-class variance, so extending it to multiple features per class with a diversity-aware sampler could further separate clusters—testable but beyond the paper.
  • The reported failure cases (false-positive 'unknown' detections) suggest the object focus loss increases recall at the cost of precision; a precision-recall trade-off study varying the entropy threshold would quantify this.
  • Since CLIP embeddings are frozen and trained on web-scale image-text data, the semantic axes may be biased toward the 80 COCO/VOC class names; evaluating on classes outside that distribution would test whether the alignment generalizes.
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

2 major / 4 minor

Summary. The paper proposes an open-set object detection (OSOD) framework built on Faster R-CNN. The contributions are: (i) a semantic clustering module that aligns proposal features with CLIP text embeddings of known classes, (ii) a class decorrelation module that imposes an orthogonality constraint on sampled features, (iii) an object focus loss combining centerness and classification-based objectness for the RPN, and (iv) an evaluation technique called entropy thresholding that relabels low-confidence predictions as unknown. The authors also introduce the Harmonic Mean Precision (HMP) metric, defined as the harmonic mean of mAPk and APu. Experiments on PASCAL VOC and MS-COCO under the VOC-COCO-T1 and T2 protocols report consistent improvements over OpenDet and other baselines, with APu gains of 17–24% on ResNet50 in the T1 setting, along with reduced WI and AOSE.

Significance. If the reported results are valid, the paper would be a useful contribution to OSOD: the semantic clustering idea is reasonable, the ablation studies are thorough, and the HMP metric, though simple, provides a convenient single-number summary. The authors also provide extensive comparisons across three backbones and multiple open-set settings, and they include failure-case analysis. However, the significance rests on the validity of the evaluation protocol. The entropy-thresholding post-processing is a potential source of asymmetric evaluation, and the threshold is tuned on the test set, which is a data-leakage concern. These issues are load-bearing because the claimed improvements over prior work are the central contribution.

major comments (2)
  1. [Supplementary B.2 and Section 4.1] The entropy threshold is set to 0.85 based on 'empirical analysis' on the VOC-COCO-40 test set (Supplementary Figure 7). Selecting a post-processing parameter on the test set and then reporting test-set results with that parameter constitutes test-set leakage, which inflates the reported performance and prevents a fair comparison with baselines. The threshold should be chosen on a held-out validation split, and the sensitivity of the main result to the threshold value should be reported. As it stands, the central claim of significant improvement is not supported by a clean evaluation protocol.
  2. [Section 3.3, Eq. (5)] The class decorrelation loss is computed on a single sampled feature per unique class per batch, and the cross-entropy with an identity matrix encourages the sampled features to be orthogonal. The paper claims this 'enhances inter-cluster distance' and 'separates the clusters,' but the loss only constrains one feature per class and does not guarantee orthogonality of all features within a cluster. The theoretical justification is therefore weak, and the small improvement observed in the ablation (Case 4 vs. Case 1 in Table 3) does not clearly establish the mechanism. Please clarify whether the constraint is intended as an approximation, and provide evidence (analytical or empirical) that it affects the entire feature distribution, or temper the claim accordingly.
minor comments (4)
  1. [Section 3.2, Eq. (2)] The denominator of the softmax uses the symbol 'n' without definition; it should be 'k' (the number of classes), consistent with Eq. (4).
  2. [References] References [19] and [20] are the same paper (Sun et al., 'Fsce: Few-shot object detection via contrastive proposal encoding'), and references [22] and [23] are also the same paper (Wu et al., 'Two-branch objectness-centric open world detection'). Please deduplicate.
  3. [Section 4.1, HMP metric] HMP is introduced as a new metric, but it is simply the harmonic mean of mAPk and APu. Please clarify whether this is intended as a contribution or as a reporting convenience, and compare it with existing combined metrics if one exists.
  4. [Implementation Details] The paper does not mention code release or provide training/evaluation scripts, which limits reproducibility. Consider adding a reproducibility statement or a link to code.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method's training objectives are grounded in external CLIP embeddings and standard losses, and the entropy-threshold concern is an evaluation-protocol issue rather than a circular derivation.

full rationale

The paper's claimed derivation chain is not circular. The semantic clustering loss (Eqs. 1-2) aligns ROI features to frozen CLIP text embeddings, an external benchmark; the class decorrelation loss (Eqs. 3-5) enforces orthogonality among per-class features; and the object focus loss (Eqs. 6-7) combines centerness with RPN objectness. None of these losses is defined in terms of the reported metrics (mAPk, APu, WI, AOSE, HMP), and none reduces to the target result by construction. The HMP metric (Eq. 10) is a post-hoc evaluation formula and is not used in training. There is no load-bearing self-citation: the only overlapping-author reference is [8], used as an external evaluation protocol rather than as justification for the proposed losses. The most legitimate concern is the entropy-thresholding post-processing in Section 4.1: the threshold 0.85 is selected on the VOC-COCO-40 test set (Supplementary B.2) and directly relabels low-confidence predictions as 'unknown', which can mechanically improve unknown-class metrics. If baselines were not given the same post-processing, the reported SOTA gains would be an evaluation fairness/validity issue. However, this is not a circular derivation: the threshold is an evaluation-time parameter, not a learned component, and the ablation study (Table 3) shows the proposed modules improve HMP under the same protocol. Thus the central method retains independent content, and the entropy-threshold concern belongs under correctness risk rather than circularity.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim rests on four free parameters (three loss weights and an evaluation threshold) and three domain or modeling assumptions. The entropy threshold is the most consequential because it directly alters the reported metrics.

free parameters (4)
  • Entropy threshold = 0.85
    Chosen by 'empirical analysis' on the VOC-COCO-40 test set (Supplementary B.2, Figure 7); used to relabel low-confidence predictions as unknown at evaluation.
  • Loss weight alpha1 (L_SC) = 0.05
    Set empirically via ablation sweeping one weight at a time (Figure 5).
  • Loss weight alpha2 (L_CD) = 0.05
    Set empirically via ablation sweeping one weight at a time (Figure 5).
  • Loss weight alpha3 (L_Obj-Focus) = 1.0
    Set empirically via ablation sweeping one weight at a time (Figure 5).
assumptions (3)
  • domain assumption CLIP text embeddings provide a semantic space where class names are meaningfully related to region proposal visual features.
    The semantic clustering loss (Eq. 2) aligns proposal features to these embeddings; if the CLIP space does not separate visually similar but semantically different classes, the proposed alignment would not help.
  • ad hoc to paper Orthogonalizing one randomly sampled feature per class in a batch enforces separation among all features of different classes.
    The class decorrelation loss (Eq. 5) applies an identity-target cross-entropy to a softmax-normalized similarity matrix of s sampled features; the paper does not prove this generalizes to the full feature distribution.
  • domain assumption COCO annotations are treated as complete when computing APu.
    The supplementary (B.2) itself notes that annotation incompleteness makes many unknown predictions count as false positives, so APu is noisy and threshold-dependent.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Open-Set Object Detection By Aligning Known Class Representations." pith.science (2026). https://pith.science/paper/RNJHSSTP

@misc{pith2026241220701,
  author       = {Pith},
  title        = {Pith review of: Open-Set Object Detection By Aligning Known Class Representations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RNJHSSTP}},
  note         = {Machine review of arXiv:2412.20701}
}
read the original abstract

Open-Set Object Detection (OSOD) has emerged as a contemporary research direction to address the detection of unknown objects. Recently, few works have achieved remarkable performance in the OSOD task by employing contrastive clustering to separate unknown classes. In contrast, we propose a new semantic clustering-based approach to facilitate a meaningful alignment of clusters in semantic space and introduce a class decorrelation module to enhance inter-cluster separation. Our approach further incorporates an object focus module to predict objectness scores, which enhances the detection of unknown objects. Further, we employ i) an evaluation technique that penalizes low-confidence outputs to mitigate the risk of misclassification of the unknown objects and ii) a new metric called HMP that combines known and unknown precision using harmonic mean. Our extensive experiments demonstrate that the proposed model achieves significant improvement on the MS-COCO & PASCAL VOC dataset for the OSOD task.

Figures

Figures reproduced from arXiv: 2412.20701 by the authors.

Figure 1
Figure 1. Effectiveness of semantic clustering, class decorrelation, [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of our proposed method. Object Focus loss: Object focus loss is a combination of LC (classification free loss) and Lobj (classification based loss). Semantic Clustering: {T1, T2, T3, . . . , Tk} represents the class embeddings of k classes and {F1, F2, F3, . . . , Fm} represents m feature embeddings. Each Fi gets aligned with its corresponding class embedding. Class Decor￾relation: We sample one feature per… view at source ↗
Figure 3
Figure 3. Demonstrates the process of object focus loss along with [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Visual comparison between our proposed model and baseline methods such as Faster R-CNN [ [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Effect of weight coefficients α1, α2 and α3 in terms of HMP measure on VOC-COCO-40 setting. sure are illustrated in [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Effect of different prompts in CLIP-based text encoder [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 8
Figure 8. Figure 8: Visual comparison between our proposed and other methods. (Zoomed-in for better view) [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: failure cases [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 27 canonical work pages

  1. [1]

    Towards open set deep networks

    Abhijit Bendale and Terrance E Boult. Towards open set deep networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1563–1572,

  2. [2]

    The overlooked elephant of object detection: Open set

    Akshay Dhamija, Manuel Gunther, Jonathan Ventura, and Terrance Boult. The overlooked elephant of object detection: Open set. In Proceedings of the IEEE/CVF Winter Confer- ence on Applications of Computer Vision, pages 1021–1030,

    work page

  3. [3]

    The pascal visual object classes (voc) challenge

    Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. The pascal visual object classes (voc) challenge. International journal of computer vision, 88:303–338, 2010. 1, 4, 5, 11

    work page 2010

  4. [4]

    Generative openmax for multi-class open set clas- sification

    ZongYuan Ge, Sergey Demyanov, Zetao Chen, and Rahil Garnavi. Generative openmax for multi-class open set clas- sification. arXiv preprint arXiv:1707.07418, 2017. 2

    arXiv 2017

  5. [5]

    Ow-detr: Open-world detection transformer

    Akshita Gupta, Sanath Narayan, KJ Joseph, Salman Khan, Fahad Shahbaz Khan, and Mubarak Shah. Ow-detr: Open-world detection transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 9235–9244, 2022. 2, 6, 11, 13

    work page 2022

  6. [6]

    Expanding low-density latent regions for open-set object detection

    Jiaming Han, Yuqiang Ren, Jian Ding, Xingjia Pan, Ke Yan, and Gui-Song Xia. Expanding low-density latent regions for open-set object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 9591–9600, 2022. 1, 2, 5, 6, 7, 11, 12, 13

    work page 2022

  7. [7]

    Momentum contrast for unsupervised visual rep- resentation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum contrast for unsupervised visual rep- resentation learning. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 9729–9738, 2020. 3

    2020

  8. [8]

    Towards open world object de- tection

    KJ Joseph, Salman Khan, Fahad Shahbaz Khan, and Vi- neeth N Balasubramanian. Towards open world object de- tection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5830–5840,

Show all 33 references
  1. [9]

    Supervised contrastive learning

    Prannay Khosla, Piotr Teterwak, Chen Wang, Aaron Sarna, Yonglong Tian, Phillip Isola, Aaron Maschinot, Ce Liu, and Dilip Krishnan. Supervised contrastive learning. Advances in neural information processing systems, 33:18661–18673,

  2. [10]

    Learning open-world object proposals without learning to classify

    Dahun Kim, Tsung-Yi Lin, Anelia Angelova, In So Kweon, and Weicheng Kuo. Learning open-world object proposals without learning to classify. IEEE Robotics and Automation Letters, 7(2):5453–5460, 2022. 2, 3, 4

    2022

  3. [11]

    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. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceeding...

    2014

  4. [12]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision, pages 10012–10022, 2021. 11, 12

    2021

  5. [13]

    Dropout sampling for robust object detection in open-set conditions

    Dimity Miller, Lachlan Nicholson, Feras Dayoub, and Niko S¨underhauf. Dropout sampling for robust object detection in open-set conditions. In 2018 IEEE International Confer- ence on Robotics and Automation (ICRA), pages 3243–3249. IEEE, 2018. 2, 6, 7, 11, 13

    2018

  6. [14]

    Open set learning with counterfac- tual images

    Lawrence Neal, Matthew Olson, Xiaoli Fern, Weng-Keen Wong, and Fuxin Li. Open set learning with counterfac- tual images. In Proceedings of the European Conference on Computer Vision (ECCV), pages 613–628, 2018. 2

    2018

  7. [15]

    C2ae: Class conditioned auto-encoder for open-set recognition

    Poojan Oza and Vishal M Patel. C2ae: Class conditioned auto-encoder for open-set recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2307–2316, 2019. 2

    2019

  8. [16]

    Spatiotempo- ral contrastive video representation learning

    Rui Qian, Tianjian Meng, Boqing Gong, Ming-Hsuan Yang, Huisheng Wang, Serge Belongie, and Yin Cui. Spatiotempo- ral contrastive video representation learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 6964–6974, 2021. 3

    2021

  9. [17]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  10. [18]

    Faster r-cnn: Towards real-time object detection with region proposal networks

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r-cnn: Towards real-time object detection with region proposal networks. In Advances in Neural Information Pro- cessing Systems, volume 28, 2015. 4, 5, 6, 7, 11, 13, 14

    2015

  11. [19]

    Fsce: Few-shot object detection via contrastive pro- posal encoding

    Bo Sun, Banghuai Li, Shengcai Cai, Ye Yuan, and Chi Zhang. Fsce: Few-shot object detection via contrastive pro- posal encoding. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 7352– 7362, 2021. 2

    2021

  12. [20]

    Fsce: Few-shot object detection via contrastive pro- posal encoding

    Bo Sun, Banghuai Li, Shengcai Cai, Ye Yuan, and Chi Zhang. Fsce: Few-shot object detection via contrastive pro- posal encoding. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 7352– 7362, 2021. 3

    2021

  13. [21]

    Fcos: Fully convolutional one-stage object detection

    Zhi Tian, Chunhua Shen, Hao Chen, and Tong He. Fcos: Fully convolutional one-stage object detection. In Proceed- ings of the IEEE/CVF international conference on computer vision, pages 9627–9636, 2019. 4, 12

    2019

  14. [22]

    Two-branch objectness-centric open world de- tection

    Yan Wu, Xiaowei Zhao, Yuqing Ma, Duorui Wang, and Xi- anglong Liu. Two-branch objectness-centric open world de- tection. In Proceedings of the 3rd International Workshop on Human-Centric Multimedia Analysis, pages 35–40, 2022. 2

    2022

  15. [23]

    Two-branch objectness-centric open world de- tection

    Yan Wu, Xiaowei Zhao, Yuqing Ma, Duorui Wang, and Xi- anglong Liu. Two-branch objectness-centric open world de- tection. In Proceedings of the 3rd International Workshop on Human-Centric Multimedia Analysis, pages 35–40, 2022. 3

    2022

  16. [24]

    Uc-owod: Unknown-classified open world object detection

    Zhiheng Wu, Yue Lu, Xingyu Chen, Zhengxing Wu, Liwen Kang, and Junzhi Yu. Uc-owod: Unknown-classified open world object detection. In European Conference on Com- puter Vision, pages 193–210. Springer, 2022. 2

    2022

  17. [25]

    Clustering- friendly representation learning via instance discrimination and feature decorrelation

    Kouta Nakata Yaling Tao, Kentaro Takagi. Clustering- friendly representation learning via instance discrimination and feature decorrelation. Proceedings of ICLR 2021, 2021. 2, 4

    2021

  18. [26]

    Classification- reconstruction learning for open-set recognition

    Ryota Yoshihashi, Wen Shao, Rei Kawakami, Shaodi You, Makoto Iida, and Takeshi Naemura. Classification- reconstruction learning for open-set recognition. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4016–4025, 2019. 2

    2019

  19. [27]

    Open-world object detection via discriminative class prototype learning

    Jinan Yu, Liyan Ma, Zhenglin Li, Yan Peng, and Shaorong Xie. Open-world object detection via discriminative class prototype learning. In 2022 IEEE International Conference on Image Processing (ICIP), pages 626–630. IEEE, 2022. 2

    2022

  20. [28]

    Contrastive learning for label efficient semantic seg- mentation

    Xiangyun Zhao, Raviteja Vemulapalli, Philip Andrew Mans- field, Boqing Gong, Bradley Green, Lior Shapira, and Ying Wu. Contrastive learning for label efficient semantic seg- mentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 10623–1063...

    2021

  21. [29]

    Learning placeholders for open-set recognition

    Da-Wei Zhou, Han-Jia Ye, and De-Chuan Zhan. Learning placeholders for open-set recognition. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4401–4410, 2021. 6, 7, 11, 12, 13, 14

    2021

  22. [30]

    Open-set object detection using classification-free object proposal and instance-level contrastive learning

    Zhongxiang Zhou, Yifei Yang, Yue Wang, and Rong Xiong. Open-set object detection using classification-free object proposal and instance-level contrastive learning. IEEE Robotics and Automation Letters, 8(3):1691–1698, 2023. 1, 2, 3, 4, 6, 7, 11, 13

    2023

  23. [31]

    Deformable detr: Deformable trans- formers for end-to-end object detection

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable detr: Deformable trans- formers for end-to-end object detection. arXiv preprint arXiv:2010.04159, 2020. 2

    2010 arXiv

  24. [32]

    Prob: Probabilistic objectness for open world object detection

    Orr Zohar, Kuan-Chieh Wang, and Serena Yeung. Prob: Probabilistic objectness for open world object detection. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 11444–11453, 2023. 2, 6, 11, 13 Supplementary Material A. Experimental Se...

    2023

  25. [33]

    Furthermore, when used with a ConvNet backbone, our proposed method im- proves the performance further and obtains significant per- formance than other methods

    This analysis reveals that the proposed method performs better when employed with a ResNet50 backbone than other methods in terms of W Iand AOSE . Furthermore, when used with a ConvNet backbone, our proposed method im- proves the performance further and obtains significant per...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.