Pith. sign in

REVIEW 4 major objections 6 minor 37 references

CLIP Unreasonable Potential in Single-Shot Face Recognition

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A frozen CLIP image encoder with single-shot text-encoder fine-tuning reaches a 20% deployment false-positive rate on a 10-person face recognition test, far below from-scratch VGG-Face and ArcFace baselines.

desk verdict Single-shot CLIP face recognition claim is undercut by a 24-image-per-person fine-tune and from-scratch baselines; honestly written but not publishable. read the letter →

arxiv 2411.12319 v2 pith:KF5BIDZ3 submitted 2024-11-19 cs.CV cs.AI

Nhan T. Luu This is my paper · ORCID
classification cs.CVcs.AI
keywords facerecognitionCLIPsingle-shotfine-tuningfalsepositiveratevision-languagepretrainingdeploymentevaluationsmall-samplelearning
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 claims that a general-purpose vision-language model, CLIP, can be turned into a face recognizer with a minimal recipe: freeze the image encoder, fine-tune only the text encoder, and supply one image per person as a 'single shot,' treating recognition as image classification with prompts such as 'This is the image of a person named ...'. On a live deployment test with ten enrolled volunteers and two strangers, the fine-tuned CLIP RN-50 achieved a false-positive rate of 20% and deployment accuracy of 75%, while VGG-Face and ArcFace models trained from scratch on the same small dataset scored 90-100% false positives and 0-16.7% deployment accuracy. The paper's central claim is that CLIP's multimodal pretraining gives it an unusual resistance to false positives in small-data deployment, without specialized facial feature extraction. If correct, this suggests a frozen general-purpose backbone plus a tiny amount of tuning can replace purpose-built face recognition pipelines for small enrollment settings.

What carries the argument

The central object is CLIP (Contrastive Language-Image Pretraining), a model that learns a shared embedding space in which images and matching text descriptions are pulled together. The carrying mechanism is the single-shot fine-tuning protocol: freeze the image encoder, backpropagate only through the text encoder, and give each enrolled person a prompt of the form 'This is the image of a person named ...' so that the model aligns that identity's text embedding with the visual features of the person's face. Recognition is then image classification over these prompt embeddings, and deployment decisions reject any prediction with confidence below 80%. The text-encoder-only gradient flow is what keeps the recipe light, while the frozen CLIP features are what the paper credits for the low false-positive rate.

What would settle it

Repeat the deployment test with the same CLIP pipeline and the 80% confidence threshold but with 20 unknown subjects instead of 2; the reported 20% FPR rests on only two strangers, so a single false acceptance moves the rate by 50 points. If CLIP's FPR on the larger stranger set is not clearly below the from-scratch baselines run under identical conditions, the central claim fails.

Watch

Extended reading notes

Core claim

Treating face recognition as a text-conditioned image classification problem, with the CLIP image encoder frozen and only the text encoder fine-tuned on roughly 240 aligned face images (one identity per class), yields a deployment false-positive rate of 20% and a false-negative rate of 50%, alongside 75% deployment accuracy. The paper's comparison models, VGG-Face and ArcFace, reach only 0-16.7% deployment accuracy and 90-100% FPR when trained from scratch on the same data under either their original settings or a generic image-classification setting. The paper also reports that prompt formulation matters little (roughly 1% variation), that an 80% confidence threshold is used to reject unrecognized faces, and that CLIP's face features are highly similar to each other (cosine similarity around 80%), which the authors identify as a remaining limitation for security-grade use.

Load-bearing premise

The comparison assumes that VGG-Face and ArcFace trained from scratch on 240 face images are meaningful baselines, even though those models are designed for large-scale training and the paper itself notes their performance collapses on small datasets.

Editorial extensions

If this is right

  • Small-enrollment face recognition can be set up with one image per person and a single frozen vision backbone, avoiding a dedicated face-feature extractor, if the 20% deployment FPR reproduces.
  • Prompt wording is not a significant tuning lever for CLIP-based face recognition; the paper measured only about 1% performance variation across its prompt variants.
  • The approach's scale is capped by text-encoder gradient memory, which grows with the number of identities; the paper had to downsize from its planned large-class experiment to 10 people.
  • Since CLIP's face embeddings sit about 80% cosine-similarity apart, the paper expects further gains from triplet sampling or additive angular margin loss rather than from better prompts.

Reading between the lines

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

  • A fairer baseline would fine-tune a face recognition model that was already pretrained on large-scale face data; the paper only compares from-scratch baselines, so part of the reported gap may reflect those models being trained outside their intended regime.
  • The single-shot text-tuning recipe is close to a linear probe on frozen CLIP features, so the 'vision-language' component may be less important than the quality of the frozen visual backbone; a frozen general-purpose image encoder with a linear classifier should be tested as a control.
  • The 20% FPR is computed from a live test with only two unknown participants, so one false acceptance changes the rate by 50 percentage points; the number is not stable enough to compare against the baselines' 90-100% without a larger stranger set.
  • If the memory cost scales with class count, a contrastive formulation or cached text embeddings could let the same idea scale to hundreds of identities; the paper leaves that extension untested.
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

4 major / 6 minor

Summary. The manuscript proposes a face-recognition pipeline based on CLIP RN-50, treating recognition as image classification with text prompts of the form "This is the image of a person named ...". The image encoder is frozen and only the text encoder is finetuned on a locally collected dataset of 10 participants (~30 images each, split 80/20). The authors report that this "single-shot finetuned" CLIP model reaches a deployment FPR of 20%, compared with 90-100% for VGG-Face and ArcFace trained from scratch, and conclude that CLIP can reduce false positives in deployment without task-specific feature extraction. The paper also includes a prompt-sensitivity experiment and a discussion of the model's high intra-person cosine similarity.

Significance. If the result were established, it would be noteworthy: a frozen CLIP visual encoder plus a lightly trained text head might provide a practical few-shot face recognition system with controllable false positives, which is relevant to security applications. The paper is transparent about its data collection and about some limitations, and the prompt-sensitivity check in Section VI.A is a useful sanity check. However, the central empirical claim is not currently supported: the baselines are trained from scratch on tiny data, the deployment evaluation is statistically fragile, and the method is not actually single-shot under the standard meaning. The contribution is therefore a suggestive pilot study at this point.

major comments (4)
  1. [III, IV.A, VII] The paper's central claim of "single-shot finetuning" is not supported by the experimental protocol. Section III states that each participant contributed approximately 30 images; Section IV.A then splits the data 80/20 for training/testing, which yields roughly 24 training images per identity. Under the standard face-recognition meaning of single-shot (one image per identity), this is few-shot, not single-shot. If "single-shot" is intended to mean a single finetuning epoch, that should be stated explicitly, but even then the experiment does not test the minimal-data regime advertised in the title, abstract, and conclusion. No result is reported with one image per identity, so the claimed minimal-data advantage is not demonstrated.
  2. [Table I, Section V] The comparison against VGG-Face and ArcFace is confounded by pretraining. CLIP RN-50 uses an image encoder pretrained on a very large image-text corpus and keeps it frozen, whereas VGG-Face and ArcFace are trained from scratch on 240 images (80% of 300). The paper itself acknowledges that these models' performance "diminishes drastically" on limited data (Section V), so their high FPR is an expected consequence of severe underfitting rather than evidence about CLIP. A fair baseline would be a pretrained face-recognition model (or a pretrained visual encoder) adapted on the same training set, and a zero-shot CLIP control should also be reported. Without such controls, Table I cannot support the conclusion that CLIP's design is responsible for the lower FPR.
  3. [IV.B, Table I] The deployment evaluation is too small to justify "significantly lower false-positive rates." The test consists of 10 enrolled and 2 unknown participants, with no per-subject trial counts or repeated trials. With only two negative subjects, a single false positive changes the FPR from 0% to 50%, so the reported 20% value has very high sampling uncertainty. The 80% confidence threshold in Section IV.B is arbitrary, and no sensitivity analysis is provided for other thresholds. The authors should report individual predictions, the number of trials per subject, confidence intervals, and a threshold sweep before claiming a significant FPR reduction.
  4. [VII] The conclusion that the method works "without the need for complex facial feature extraction techniques" overstates the result: the pipeline still relies on CLIP's large pretrained image encoder to extract visual features, and only the text encoder is finetuned. The valid statement is that the method avoids task-specific feature-extraction training, not that feature extraction is eliminated. More importantly, the broader generalization claim that CLIP "offers a promising solution" is not supported by a 10-class, 12-subject deployment study, especially given the scalability limitation disclosed in Section VI.C.
minor comments (6)
  1. [Abstract] The abstract contains grammatical errors (e.g., "This integration demonstrating") and informal phrasing (e.g., "without the need of mass facial features extraction"); the manuscript should be carefully proofread.
  2. [Section I] The sentence beginning "One of the big problem distinct facial recognition with other classification task..." is ungrammatical and should be rewritten for clarity.
  3. [Equation (1)] The notation in the loss equation is unclear: "mini-classes" is not defined, and the summation indices are ambiguous. The softmax cross-entropy formula should be written cleanly with all quantities defined.
  4. [IV.A] The number of finetuning epochs is not reported. If "single-shot" refers to one epoch, this should be stated explicitly; otherwise the training budget should be listed.
  5. [VI.A] The prompt-sensitivity result is reported only as "approximately 1%" difference; exact accuracy values or a small table should be included so that the claim can be checked.
  6. [VI.B] The statement that CLIP's face features have "approximately 80%" cosine similarity should report the actual distribution or range rather than a single approximate value.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the empirical comparison rests on external pretrained features, explicit metric definitions, and openly acknowledged limitations; the 'single-shot' terminology issue is a correctness/fairness problem, not circularity.

full rationale

I walked the paper's claimed derivation chain and found no load-bearing step that reduces to its own inputs. The only quantitative outputs are deployment accuracy, FPR, and FNR computed from Eqs. (2)-(4), which are standard definitions from raw TP/TN/FP/FN counts rather than quantities fitted from the conclusion. The 80% confidence cutoff in Section IV.B is an explicitly stated evaluation threshold, not a parameter derived from the result. The paper's central claim about CLIP rests on its pretrained vision-language features (an external, verifiable artifact) plus a finetuning stage that uses around 24 images per identity from the 80/20 split described in Sections III and IV.A; whether this deserves the label 'single-shot' is a terminology and experimental-design concern, not a circular one, because the reported FPR is not constructed from the meaning of the phrase. The comparison with VGG-Face and ArcFace trained from scratch on 240 images is confounded, and Section V itself concedes that these models' performance 'diminishes drastically' in that regime, but an unfair baseline is an external-validity problem rather than a demonstration that the CLIP result is equivalent to the comparison inputs. The only self-citation, ref. [24], appears in Section II.B as one item in a list of CLIP-improvement works and is not load-bearing for any experimental conclusion. Section VI.C candidly acknowledges the GPU-memory limitation that forced the 10-class scale, which is a scope limitation rather than a circular step. Accordingly, no predicted quantity is forced by definition, by fitted inputs, or by a self-citation chain, and the paper should not receive a circularity score above zero.

Assumptions & free parameters 1 free parameters · 2 assumptions · 0 invented entities

The paper contains no mathematical derivation, so the only free parameter is the evaluation confidence threshold. The main axioms are domain assumptions about CLIP's pretrained features and the representativeness of the deployment test.

free parameters (1)
  • confidence threshold = 80%
    Predictions with confidence below 80% are counted as unrecognized in Equations (2)-(4). This threshold is chosen without sensitivity analysis and directly controls FPR and FNR.
assumptions (2)
  • domain assumption CLIP's frozen image encoder provides features that are informative for face recognition after text-encoder fine-tuning
    The method freezes the image encoder (Section IV.A), relying on pretrained CLIP features to separate identities.
  • domain assumption The deployment environment matches the training data distribution closely enough that same-device and same-camera conditions are representative
    Section IV.B states the deployment test uses the same device and camera setup to ensure environmental consistency; this is assumed to be valid for evaluating FPR.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CLIP Unreasonable Potential in Single-Shot Face Recognition." pith.science (2026). https://pith.science/paper/KF5BIDZ3

@misc{pith2026241112319,
  author       = {Pith},
  title        = {Pith review of: CLIP Unreasonable Potential in Single-Shot Face Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KF5BIDZ3}},
  note         = {Machine review of arXiv:2411.12319}
}
read the original abstract

Face recognition is a core task in computer vision designed to identify and authenticate individuals by analyzing facial patterns and features. This field intersects with artificial intelligence image processing and machine learning with applications in security authentication and personalization. Traditional approaches in facial recognition focus on capturing facial features like the eyes, nose and mouth and matching these against a database to verify identities. However challenges such as high false positive rates have persisted often due to the similarity among individuals facial features. Recently Contrastive Language Image Pretraining (CLIP) a model developed by OpenAI has shown promising advancements by linking natural language processing with vision tasks allowing it to generalize across modalities. Using CLIP's vision language correspondence and single-shot finetuning the model can achieve lower false positive rates upon deployment without the need of mass facial features extraction. This integration demonstrating CLIP's potential to address persistent issues in face recognition model performance without complicating our training paradigm.

Figures

Figures reproduced from arXiv: 2411.12319 by the authors.

Figure 1
Figure 1. A diagram illustrate our dataset acquisition and preprocessing method. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. A graph comparing traditional face recognition pipeline and our method using single-shot finetunned CLIP model. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 13 canonical work pages

  1. [27]

    Face recognition in the age of clip & billion image datasets,

    A. Bhat and S. Jain, “Face recognition in the age of clip & billion image datasets,” arXiv preprint arXiv:2301.07315 , 2023

    arXiv 2023

  2. [28]

    Clip-cluster: Clip-guided attribute hallucination for face clustering,

    S. Shen, W. Li, X. Wang, D. Zhang, Z. Jin, J. Zhou, and J. Lu, “Clip-cluster: Clip-guided attribute hallucination for face clustering,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 20 786–20 795

    2023

  3. [1]

    A survey of face recognition,

    X. Wang, J. Peng, S. Zhang, B. Chen, Y . Wang, and Y . Guo, “A survey of face recognition,” arXiv preprint arXiv:2212.13038 , 2022

    arXiv 2022

  4. [2]

    Face recognition systems: A survey,

    Y . Kortli, M. Jridi, A. Al Falou, and M. Atri, “Face recognition systems: A survey,” Sensors, vol. 20, no. 2, p. 342, 2020

    work page 2020

  5. [3]

    Deep face recognition: A survey,

    M. Wang and W. Deng, “Deep face recognition: A survey,” Neurocom- puting, vol. 429, pp. 215–244, 2021

    work page 2021

  6. [4]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al., “Learning transferable visual models from natural language supervision,” in International conference on machine learning . PMLR, 2021, pp. 8748–8763

    2021

  7. [5]

    Face recognition using eigenfaces,

    M. A. Turk and A. P. Pentland, “Face recognition using eigenfaces,” in Proceedings. 1991 IEEE computer society conference on computer vision and pattern recognition . IEEE Computer Society, 1991, pp. 586–587

    work page 1991

  8. [6]

    Face description with local binary patterns: Application to face recognition,

    T. Ahonen, A. Hadid, and M. Pietikainen, “Face description with local binary patterns: Application to face recognition,” IEEE transactions on pattern analysis and machine intelligence , vol. 28, no. 12, pp. 2037– 2041, 2006

    2006

Show all 37 references
  1. [7]

    Eigenfaces vs. fisherfaces: Recognition using class specific linear projection,

    P. N. Belhumeur, J. P. Hespanha, and D. J. Kriegman, “Eigenfaces vs. fisherfaces: Recognition using class specific linear projection,” IEEE Transactions on pattern analysis and machine intelligence, vol. 19, no. 7, pp. 711–720, 1997

    1997

  2. [8]

    Deepface: Closing the gap to human-level performance in face verification,

    Y . Taigman, M. Yang, M. Ranzato, and L. Wolf, “Deepface: Closing the gap to human-level performance in face verification,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2014, pp. 1701–1708

    2014

  3. [9]

    Facenet: A unified embed- ding for face recognition and clustering,

    F. Schroff, D. Kalenichenko, and J. Philbin, “Facenet: A unified embed- ding for face recognition and clustering,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2015, pp. 815– 823

    2015

  4. [10]

    Deep face recognition,

    O. Parkhi, A. Vedaldi, and A. Zisserman, “Deep face recognition,” in BMVC 2015-Proceedings of the British Machine Vision Conference

    2015

  5. [11]

    Cosface: Large margin cosine loss for deep face recognition,

    H. Wang, Y . Wang, Z. Zhou, X. Ji, D. Gong, J. Zhou, Z. Li, and W. Liu, “Cosface: Large margin cosine loss for deep face recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 5265–5274

    2018

  6. [12]

    Zegclip: Towards adapting clip for zero-shot semantic segmentation,

    Z. Zhou, Y . Lei, B. Zhang, L. Liu, and Y . Liu, “Zegclip: Towards adapting clip for zero-shot semantic segmentation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 11 175–11 185

    2023

  7. [13]

    Clipn for zero-shot ood detection: Teaching clip to say no,

    H. Wang, Y . Li, H. Yao, and X. Li, “Clipn for zero-shot ood detection: Teaching clip to say no,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 1802–1812

    2023

  8. [14]

    Clip-forge: Towards zero-shot text-to-shape generation,

    A. Sanghi, H. Chu, J. G. Lambourne, Y . Wang, C.-Y . Cheng, M. Fumero, and K. R. Malekshan, “Clip-forge: Towards zero-shot text-to-shape generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 18 603–18 613

    2022

  9. [15]

    Clip the gap: A single domain generalization approach for object detection,

    V . Vidit, M. Engilberge, and M. Salzmann, “Clip the gap: A single domain generalization approach for object detection,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 3219–3229

    2023

  10. [16]

    Open-vocabulary object detection via vision and language knowledge distillation,

    X. Gu, T.-Y . Lin, W. Kuo, and Y . Cui, “Open-vocabulary object detection via vision and language knowledge distillation,” arXiv preprint arXiv:2104.13921, 2021

    2021 arXiv

  11. [17]

    Hierarchical text-conditional image generation with clip latents,

    A. Ramesh, P. Dhariwal, A. Nichol, C. Chu, and M. Chen, “Hierarchical text-conditional image generation with clip latents,” arXiv preprint arXiv:2204.06125, vol. 1, no. 2, p. 3, 2022

    2022 arXiv

  12. [18]

    Styleclip: Text-driven manipulation of stylegan imagery,

    O. Patashnik, Z. Wu, E. Shechtman, D. Cohen-Or, and D. Lischinski, “Styleclip: Text-driven manipulation of stylegan imagery,” in Proceed- ings of the IEEE/CVF international conference on computer vision , 2021, pp. 2085–2094

    2021

  13. [19]

    Learning to prompt for vision- language models,

    K. Zhou, J. Yang, C. C. Loy, and Z. Liu, “Learning to prompt for vision- language models,” International Journal of Computer Vision , vol. 130, no. 9, pp. 2337–2348, 2022

    2022

  14. [20]

    Galip: Generative adversarial clips for text-to-image synthesis,

    M. Tao, B.-K. Bao, H. Tang, and C. Xu, “Galip: Generative adversarial clips for text-to-image synthesis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 14 214–14 223

    2023

  15. [21]

    Clip-gen: Language- free training of a text-to-image generator with clip,

    Z. Wang, W. Liu, Q. He, X. Wu, and Z. Yi, “Clip-gen: Language- free training of a text-to-image generator with clip,” arXiv preprint arXiv:2203.00386, 2022

    2022 arXiv

  16. [22]

    The unreasonable effectiveness of clip features for image captioning: an experimental analysis,

    M. Barraco, M. Cornia, S. Cascianelli, L. Baraldi, and R. Cucchiara, “The unreasonable effectiveness of clip features for image captioning: an experimental analysis,” in proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2022, pp. 4662–4670

    2022

  17. [23]

    Clipcap: Clip prefix for image captioning,

    R. Mokady, A. Hertz, and A. H. Bermano, “Clipcap: Clip prefix for image captioning,” arXiv preprint arXiv:2111.09734 , 2021

    2021 arXiv

  18. [24]

    Blind image quality assessment with multimodal prompt learning,

    N. T. Luu, C. Onuoha, and T. C. Thang, “Blind image quality assessment with multimodal prompt learning,” in2023 IEEE 15th International Con- ference on Computational Intelligence and Communication Networks (CICN). IEEE, 2023, pp. 614–618

    2023

  19. [25]

    Maple: Multi-modal prompt learning,

    M. U. Khattak, H. Rasheed, M. Maaz, S. Khan, and F. S. Khan, “Maple: Multi-modal prompt learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 19 113–19 122

    2023

  20. [26]

    Conditional prompt learning for vision-language models,

    K. Zhou, J. Yang, C. C. Loy, and Z. Liu, “Conditional prompt learning for vision-language models,” in Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition , 2022, pp. 16 816– 16 825

    2022

  21. [29]

    Sample and com- putation redistribution for efficient face detection,

    J. Guo, J. Deng, A. Lattas, and S. Zafeiriou, “Sample and com- putation redistribution for efficient face detection,” arXiv preprint arXiv:2105.04714, 2021

    2021 arXiv

  22. [30]

    Exploring clip for assessing the look and feel of images,

    J. Wang, K. C. Chan, and C. C. Loy, “Exploring clip for assessing the look and feel of images,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 2, 2023, pp. 2555–2563

    2023

  23. [31]

    Blind image quality assessment via vision-language correspondence: A multitask learning perspective,

    W. Zhang, G. Zhai, Y . Wei, X. Yang, and K. Ma, “Blind image quality assessment via vision-language correspondence: A multitask learning perspective,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 14 071–14 081

    2023

  24. [32]

    Decoupled weight decay regularization,

    I. Loshchilov, “Decoupled weight decay regularization,” arXiv preprint arXiv:1711.05101, 2017

    2017 arXiv

  25. [33]

    Stochastic gradient descent with warm restarts,

    I. Loshchilov and F. Hutter, “Stochastic gradient descent with warm restarts,” in Proceedings of the 5th International Conference on Learning Representations, pp. 1–16

  26. [34]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al. , “Pytorch: An imperative style, high-performance deep learning library,” Advances in neural information processing systems , vol. 32, 2019

    2019

  27. [35]

    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,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4690– 4699

    2019

  28. [36]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980 , 2014

    2014 arXiv

  29. [2015]

    British Machine Vision Association, 2015

    2015

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.