完整搜尋頁面 →

並排比較邊界損失:一個公式與它的特殊情況

選讀・DLCV 02A・臉部辨識與度量學習 的補充資料(第 3/6 則)。主線不依賴這則補充;引用編號對應本頁底部的參考文獻。

導覽:本頁與本系列: 導覽:本頁與本系列

本頁內容

補充資料 · 6

  1. 對齊臉部:最小平方相似變換
  2. 尺度 s 要多大?
  3. 並排比較邊界損失:一個公式與它的特殊情況
  4. 一個類別不只一個中心:子中心、品質感知邊界與 Partial FC
  5. 量測極小的錯誤率
  6. 深度學習之前:Eigenfaces 與 Fisherfaces

論文導讀 · 3

  1. ArcFace(Deng 等人,CVPR 2019;TPAMI 2022)
  2. NIST FRVT Part 3:族群效應(Grother、Ngan 與 Hanaoka,2019)
  3. 現代觀點綜述導讀

電腦視覺深度學習

  1. 02 · 深度影像分類
  2. 03 · 物件偵測:從 R-CNN 到 DETR
  3. 09 · 深度影像復原:低階視覺
查看完整課程樹 →

這則補充把三種邊界寫成同一個公式,推導它們的雙類別邊界,並給出玩具程式碼所用的梯度。

一個公式。 ArcFace 的期刊論文把 SphereFace、ArcFace 與 CosFace 寫成組合邊界目標 logit 的特例 [16]

ψ(θ)=cos⁡(m1θ+m2)−m3,\psi(\theta) = \cos(m_1 \theta + m_2) - m_3,

其中 m1m_1 是 SphereFace 的乘法式角度邊界,m2m_2 是 ArcFace 的加法式角度邊界,m3m_3 是 CosFace 的加法式餘弦邊界。(m1,m2,m3)=(1,0,0)(m_1, m_2, m_3) = (1, 0, 0) 是正規化 softmax;(1,0.5,0)(1, 0.5, 0) 是 ArcFace;(1,0,0.35)(1, 0, 0.35) 是 CosFace;論文也嘗試了三者的組合 [16]。

雙類別邊界。 令中心 W1W_1 與 W2W_2 相距角度 φ\varphi,並考慮落在它們之間弧上的特徵,它與 W1W_1 的夾角為 θ1\theta_1,與 W2W_2 的夾角為 θ2=φ−θ1\theta_2 = \varphi - \theta_1。當 ψ(θ1)>cos⁡θ2\psi(\theta_1) \gt \cos\theta_2 時,類別 1 被滿足。

  • ArcFace: cos⁡(θ1+m)>cos⁡(φ−θ1)\cos(\theta_1 + m) \gt \cos(\varphi - \theta_1)。兩個引數都落在 [0,π][0, \pi] 內(當 θ1+m≤π\theta_1 + m \le \pi 時),餘弦在那裡遞減,所以條件是 θ1+m<φ−θ1\theta_1 + m \lt \varphi - \theta_1,也就是 θ1<(φ−m)/2\theta_1 \lt (\varphi - m)/2。由對稱性,類別 2 在 θ1>(φ+m)/2\theta_1 \gt (\varphi + m)/2 時被滿足,所以兩者都不滿足的區域寬度恰好是 mm,與 φ\varphi 無關。
  • SphereFace(在 ψ\psi 的第一段上):mθ1<φ−θ1m\theta_1 \lt \varphi - \theta_1,所以 θ1<φ/(m+1)\theta_1 \lt \varphi/(m+1);間隙是 φ−2φ/(m+1)=φ(m−1)/(m+1)\varphi - 2\varphi/(m+1) = \varphi(m-1)/(m+1)。中心靠近時邊界小,中心遠離時邊界大。
  • CosFace: cos⁡θ1−cos⁡(φ−θ1)>m\cos\theta_1 - \cos(\varphi - \theta_1) \gt m。利用 cos⁡A−cos⁡B=−2sin⁡A+B2sin⁡A−B2\cos A - \cos B = -2\sin\frac{A+B}{2}\sin\frac{A-B}{2},這等於 2sin⁡φ2sin⁡(φ2−θ1)>m2\sin\frac{\varphi}{2}\sin\big(\frac{\varphi}{2} - \theta_1\big) \gt m,所以 θ1<φ2−arcsin⁡ ⁣(m2sin⁡(φ/2))\theta_1 \lt \frac{\varphi}{2} - \arcsin\!\big(\frac{m}{2\sin(\varphi/2)}\big)。角度間隙為 2arcsin⁡ ⁣(m2sin⁡(φ/2))2\arcsin\!\big(\frac{m}{2\sin(\varphi/2)}\big):中心靠近時較大,而若 2sin⁡(φ/2)<m2\sin(\varphi/2) \lt m 則沒有定義(弧上沒有可滿足的區域)。當 φ=80∘\varphi = 80^\circ 且 m=0.35m = 0.35 時約為 31∘31^\circ,如圖 2。

ArcFace logit 的梯度。 訓練使用目標 logit 對餘弦 c=cos⁡θc = \cos\theta 的導數,因為網路計算的是餘弦。令 θ=arccos⁡c\theta = \arccos c,

∂∂c scos⁡(θ+m)=s sin⁡(θ+m) 1sin⁡θ,\frac{\partial}{\partial c}\, s\cos(\theta + m) = s\,\sin(\theta + m)\,\frac{1}{\sin\theta},

因為 ddθcos⁡(θ+m)=−sin⁡(θ+m)\frac{d}{d\theta}\cos(\theta + m) = -\sin(\theta + m) 且 dθdc=−1/sin⁡θ\frac{d\theta}{dc} = -1/\sin\theta。這就是玩具程式碼中的 np.sin(th + margin) / np.sin(th) 因子。只要 sin⁡(θ+m)>sin⁡θ\sin(\theta + m) \gt \sin\theta,也就是 θ<(π−m)/2\theta \lt (\pi - m)/2,它就大於 1,所以邊界會放大那些已經靠近中心的樣本的梯度,直到 softmax 機率飽和。由此還有兩個特殊情況。在 θ=0\theta = 0 附近,這個因子會除以 sin⁡θ≈0\sin\theta \approx 0,所以實作會把餘弦裁切到離 ±1\pm 1 一點距離的地方(玩具程式碼裁切在 1−10−71 - 10^{-7})。當 θ+m>π\theta + m \gt \pi,這個因子變成負的,logit 不再單調,如正文所述;這樣的樣本需要備用的懲罰。

為什麼容易的樣本會停止貢獻。 對 logits 的梯度是 pj−[j=yi]p_j - [j = y_i],其中 pjp_j 是 softmax 機率。當 s ψ(θyi)s\,\psi(\theta_{y_i}) 遠高於每個 scos⁡θjs\cos\theta_j 時,pyi→1p_{y_i} \to 1,這些差值全都趨近於零。邊界移動了這件事發生的位置:樣本必須進入自己類別的滿足區域一段明確的距離,它的梯度才會消失,這正是邊界產生緊密叢集的方式。

參考文獻

  1. M. Wang and W. Deng, “Deep face recognition: A survey,” Neurocomputing, vol. 429, pp. 215–244, 2021. doi
  2. P. Grother, M. Ngan and K. Hanaoka, “Face Recognition Vendor Test (FRVT) Part 3: Demographic Effects,” NIST Interagency Report 8280, National Institute of Standards and Technology, 2019. doi
  3. National Institute of Standards and Technology, “Face Recognition Technology Evaluation (FRTE) 1:1 Verification,” evaluation web page, accessed 2026. NIST
  4. K. Zhang, Z. Zhang, Z. Li and Y. Qiao, “Joint face detection and alignment using multitask cascaded convolutional networks,” IEEE Signal Processing Letters, vol. 23, no. 10, pp. 1499–1503, 2016. doi
  5. J. Deng, J. Guo, E. Ververas, I. Kotsia and S. Zafeiriou, “RetinaFace: Single-shot multi-level face localisation in the wild,” in Proc. IEEE/CVF CVPR, 2020. CVF
  6. Y. Taigman, M. Yang, M. Ranzato and L. Wolf, “DeepFace: Closing the gap to human-level performance in face verification,” in Proc. IEEE CVPR, 2014. CVF
  7. F. Schroff, D. Kalenichenko and J. Philbin, “FaceNet: A unified embedding for face recognition and clustering,” in Proc. IEEE CVPR, 2015. CVF
  8. S. Chopra, R. Hadsell and Y. LeCun, “Learning a similarity metric discriminatively, with application to face verification,” in Proc. IEEE CVPR, vol. 1, pp. 539–546, 2005. doi
  9. R. Hadsell, S. Chopra and Y. LeCun, “Dimensionality reduction by learning an invariant mapping,” in Proc. IEEE CVPR, vol. 2, pp. 1735–1742, 2006. doi
  10. A. Hermans, L. Beyer and B. Leibe, “In defense of the triplet loss for person re-identification,” arXiv:1703.07737, 2017. arXiv
  11. F. Wang, X. Xiang, J. Cheng and A. L. Yuille, “NormFace: L2 hypersphere embedding for face verification,” in Proc. ACM International Conference on Multimedia, pp. 1041–1049, 2017. doi
  12. W. Liu, Y. Wen, Z. Yu, M. Li, B. Raj and L. Song, “SphereFace: Deep hypersphere embedding for face recognition,” in Proc. IEEE CVPR, 2017. CVF
  13. 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 Proc. IEEE/CVF CVPR, 2018. CVF
  14. F. Wang, J. Cheng, W. Liu and H. Liu, “Additive margin softmax for face verification,” IEEE Signal Processing Letters, vol. 25, no. 7, pp. 926–930, 2018. doi
  15. J. Deng, J. Guo, N. Xue and S. Zafeiriou, “ArcFace: Additive angular margin loss for deep face recognition,” in Proc. IEEE/CVF CVPR, 2019. CVF
  16. J. Deng, J. Guo, J. Yang, N. Xue, I. Kotsia and S. Zafeiriou, “ArcFace: Additive angular margin loss for deep face recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 10, pp. 5962–5979, 2022. doi
  17. Y. Wen, K. Zhang, Z. Li and Y. Qiao, “A discriminative feature learning approach for deep face recognition,” in Proc. ECCV, pp. 499–515, 2016. doi
  18. J. Deng, J. Guo, T. Liu, M. Gong and S. Zafeiriou, “Sub-center ArcFace: Boosting face recognition by large-scale noisy web faces,” in Proc. ECCV, pp. 741–757, 2020. doi
  19. Q. Meng, S. Zhao, Z. Huang and F. Zhou, “MagFace: A universal representation for face recognition and quality assessment,” in Proc. IEEE/CVF CVPR, 2021. CVF
  20. M. Kim, A. K. Jain and X. Liu, “AdaFace: Quality adaptive margin for face recognition,” in Proc. IEEE/CVF CVPR, 2022. CVF
  21. X. An, J. Deng, J. Guo, Z. Feng, X. Zhu, J. Yang and T. Liu, “Killing two birds with one stone: Efficient and robust training of face recognition CNNs by Partial FC,” in Proc. IEEE/CVF CVPR, 2022. CVF
  22. A. Martin, G. Doddington, T. Kamm, M. Ordowski and M. Przybocki, “The DET curve in assessment of detection task performance,” in Proc. Eurospeech, pp. 1895–1898, 1997. ISCA
  23. G. B. Huang, M. Mattar, T. Berg and E. Learned-Miller, “Labeled Faces in the Wild: A database for studying face recognition in unconstrained environments,” in Workshop on Faces in ‘Real-Life’ Images: Detection, Alignment, and Recognition, Marseille, 2008. HAL
  24. I. Kemelmacher-Shlizerman, S. M. Seitz, D. Miller and E. Brossard, “The MegaFace benchmark: 1 million faces for recognition at scale,” in Proc. IEEE CVPR, 2016. CVF
  25. University of Washington, “MegaFace and MF2: Million-scale face recognition,” project web page, accessed 2026. project page
  26. Y. Guo, L. Zhang, Y. Hu, X. He and J. Gao, “MS-Celeb-1M: A dataset and benchmark for large-scale face recognition,” in Proc. ECCV, pp. 87–102, 2016. doi
  27. C. Whitelam, E. Taborsky, A. Blanton, B. Maze, J. Adams, T. Miller, N. Kalka, A. K. Jain, J. A. Duncan, K. Allen, J. Cheney and P. Grother, “IARPA Janus Benchmark-B face dataset,” in Proc. IEEE CVPR Workshops, 2017. CVF
  28. B. Maze, J. Adams, J. A. Duncan, N. Kalka, T. Miller, C. Otto, A. K. Jain, W. T. Niggel, J. Anderson, J. Cheney and P. Grother, “IARPA Janus Benchmark – C: Face dataset and protocol,” in Proc. International Conference on Biometrics (ICB), pp. 158–165, 2018. doi
  29. Z. Zhu, G. Huang, J. Deng, Y. Ye, J. Huang, X. Chen, J. Zhu, T. Yang, J. Lu, D. Du and J. Zhou, “WebFace260M: A benchmark unveiling the power of million-scale deep face recognition,” in Proc. IEEE/CVF CVPR, 2021. CVF
  30. K. Peng, A. Mathur and A. Narayanan, “Mitigating dataset harms requires stewardship: Lessons from 1000 papers,” in Proc. NeurIPS Datasets and Benchmarks Track, 2021. proceedings
  31. G. Bae, M. de La Gorce, T. Baltrušaitis, C. Hewitt, D. Chen, J. Valentin, R. Cipolla and J. Shen, “DigiFace-1M: 1 million digital face images for face recognition,” in Proc. IEEE/CVF WACV, 2023. CVF
  32. M. Kim, F. Liu, A. Jain and X. Liu, “DCFace: Synthetic face generation with dual condition diffusion model,” in Proc. IEEE/CVF CVPR, 2023. CVF
  33. F. Boutros, V. Struc, J. Fierrez and N. Damer, “Synthetic data for face recognition: Current state and future prospects,” Image and Vision Computing, vol. 135, art. 104688, 2023. doi
  34. M. Wang, W. Deng, J. Hu, X. Tao and Y. Huang, “Racial Faces in the Wild: Reducing racial bias by information maximization adaptation network,” in Proc. IEEE/CVF ICCV, pp. 692–702, 2019. doi
  35. J. Buolamwini and T. Gebru, “Gender Shades: Intersectional accuracy disparities in commercial gender classification,” in Proc. Conference on Fairness, Accountability and Transparency, PMLR 81, pp. 77–91, 2018. PMLR
  36. P. Drozdowski, C. Rathgeb, A. Dantcheva, N. Damer and C. Busch, “Demographic bias in biometrics: A survey on an emerging challenge,” IEEE Transactions on Technology and Society, vol. 1, no. 2, pp. 89–103, 2020. doi
  37. Z. Yu, Y. Qin, X. Li, C. Zhao, Z. Lei and G. Zhao, “Deep learning for face anti-spoofing: A survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, pp. 5609–5631, 2023. doi
  38. European Commission, “What personal data is considered sensitive?,” data protection guidance referring to Article 9 of the GDPR, accessed 2026. European Commission
  39. European Commission, “AI Act” (Regulation (EU) 2024/1689), policy page, accessed 2026. European Commission
  40. M. Turk and A. Pentland, “Eigenfaces for recognition,” Journal of Cognitive Neuroscience, vol. 3, no. 1, pp. 71–86, 1991. doi
  41. 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. doi