Pith. sign in

REVIEW 6 cited by

ArcFace: Additive Angular Margin Loss for Deep Face Recognition

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1801.07698 v4 pith:3YPHP27X submitted 2018-01-23 cs.CV

classification cs.CV
keywords arcfacefacediscriminativefacesfeaturelossonlytraining
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

Recently, a popular line of research in face recognition is adopting margins in the well-established softmax loss function to maximize class separability. In this paper, we first introduce an Additive Angular Margin Loss (ArcFace), which not only has a clear geometric interpretation but also significantly enhances the discriminative power. Since ArcFace is susceptible to the massive label noise, we further propose sub-center ArcFace, in which each class contains $K$ sub-centers and training samples only need to be close to any of the $K$ positive sub-centers. Sub-center ArcFace encourages one dominant sub-class that contains the majority of clean faces and non-dominant sub-classes that include hard or noisy faces. Based on this self-propelled isolation, we boost the performance through automatically purifying raw web faces under massive real-world noise. Besides discriminative feature embedding, we also explore the inverse problem, mapping feature vectors to face images. Without training any additional generator or discriminator, the pre-trained ArcFace model can generate identity-preserved face images for both subjects inside and outside the training data only by using the network gradient and Batch Normalization (BN) priors. Extensive experiments demonstrate that ArcFace can enhance the discriminative feature embedding as well as strengthen the generative face synthesis.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

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

  1. DriveDNA: A Large-Scale Multimodal Naturalistic Driving Dataset and Benchmark for Driving Style Identification

    cs.LG 2026-07 conditional novelty 7.0 of 10

    A multi-vehicle naturalistic benchmark finds learned driving embeddings retain driver identity under condition matching, while descriptors collapse and video re-ID is mostly route leakage.

  2. Adaptive Identity Anchoring: Closed-Loop Keyframe Placement for Synthetic Paired Supervision in Video Face Swapping

    cs.CV 2026-07 conditional novelty 6.0 of 10

    For video face swapping, adaptively adding swapped anchor frames at the moments of worst identity drift should make synthetic training pairs more faithful than the current first-and-last-frame-only scheme.

  3. SoftReason: A Fully Differentiable Neuro-Soft-Symbolic Deductive Reasoning Architecture over High-Dimensional Perceptual Data

    cs.AI 2026-07 conditional novelty 6.0 of 10

    SoftReason learns a differentiable soft deductive closure operator over perceptual facts and reports 94.3% Hit@1 on KVQA entity linking.

  4. FaceGCD: Generalized Face Discovery via Dynamic Prefix Generation

    cs.CV 2025-07 conditional novelty 6.0 of 10

    FaceGCD uses a hypernetwork to generate instance-specific layer-wise prefixes, achieving state-of-the-art accuracy on new generalized face discovery benchmarks built from YouTube Faces and CASIA-WebFaces.

  5. A Responsible Face Recognition Approach for Small and Mid-Scale Systems Through Personalized Neural Networks

    cs.CV 2025-05 conditional novelty 6.0 of 10

    MOTE replaces fixed face embeddings with per-identity binary classifiers trained using KDE-generated synthetic samples, improving gender fairness and privacy at the cost of storage and enrollment time.

  6. TaoMate: Anchor-Guided Memory Bridging Evolving and Reference States for Real-Time Audio-Video Digital Human Generation

    cs.CV 2026-07 conditional novelty 5.0 of 10

    Anchor-guided fixed-capacity multimodal memory plus stage-parallel few-step denoising stabilizes long-form causal audio-video digital humans at real-time speed.

Pith tools