Learning Steadily: Accumulating Relative Point Margin Scores for Face Image Quality Assessment

Hugging Face Daily Papers Papers
face-quality-assessment face-recognition biometrics training-stability cumulative-averaging deep-learning

Summary

该论文提出CARPM-FIQA策略，通过在整个训练轨迹上累积相对点边际分数（类内紧凑性与类间分离度之比）来解决FR集成式人脸图像质量评估中的训练不稳定问题，并在多个基准测试上验证了其有效性。

Face Image Quality Assessment determines the suitability of captured face images for automated face recognition (FR), a critical capability for reliable biometric systems. Existing state-of-the-art FR-integrated FIQA methods suffer from temporal instability: as the feature space evolves during training, single-epoch quality estimates fluctuate, creating a moving target that undermines reliable quality prediction. We introduce CARPM-FIQA, a stabilization strategy for FR-integrated FIQA that accumulates relative point margin measurements, the ratio between intra-class compactness and inter-class separation, across the entire training trajectory rather than relying on single-epoch estimates. This cumulative averaging approach provides theoretically grounded advantages: reduced variance in quality estimates, improved mean squared error, and enhanced ranking stability with convergence guarantees as training progresses. Through controlled experiments on the SynFIQA dataset with labeled quality groups, we demonstrate that cumulative averaging achieves superior discriminative ability, and ablation studies across different training configurations confirm consistent improvements. Evaluated against twelve FIQA methods on eight challenging benchmarks with four FR models at two FMR thresholds, CARPM-FIQA places 4th (CARPM-FIQA(L)) and 6th (CARPM-FIQA(S)) of 17 compared methods by pAUC-EDC and AUC-EDC averaged across FR models and, after per-benchmark normalization, across benchmarks, staying within a few percent of the best method's normalized average for every FR model, providing a principled solution to training instability while maintaining the performance benefits of FR integration. More broadly, our work demonstrates that temporal aggregation strategies can stabilize training objectives in deep learning systems where target values inherently fluctuate due to evolving feature representations.
Original Article
View Cached Full Text

Cached at: 10/06/26, 12:55 PM

Paper page - Learning Steadily: Accumulating Relative Point Margin Scores for Face Image Quality Assessment

Source: https://huggingface.co/papers/2609.31662

Abstract

FaceImageQualityAssessmentdeterminesthesuitabilityofcapturedfaceimagesforautomatedfacerecognition(FR),acriticalcapabilityforreliablebiometricsystems.Existingstate-of-the-artFR-integratedFIQAmethodssufferfromtemporalinstability:asthefeaturespaceevolvesduringtraining,single-epochqualityestimatesfluctuate,creatingamovingtargetthatunderminesreliablequalityprediction.WeintroduceCARPM-FIQA,astabilizationstrategyforFR-integratedFIQAthataccumulatesrelativepointmarginmeasurements,theratiobetweenintra-classcompactnessandinter-classseparation,acrosstheentiretrainingtrajectoryratherthanrelyingonsingle-epochestimates.Thiscumulativeaveragingapproachprovidestheoreticallygroundedadvantages:reducedvarianceinqualityestimates,improvedmeansquarederror,andenhancedrankingstabilitywithconvergenceguaranteesastrainingprogresses.ThroughcontrolledexperimentsontheSynFIQAdatasetwithlabeledqualitygroups,wedemonstratethatcumulativeaveragingachievessuperiordiscriminativeability,andablationstudiesacrossdifferenttrainingconfigurationsconfirmconsistentimprovements.EvaluatedagainsttwelveFIQAmethodsoneightchallengingbenchmarkswithfourFRmodelsattwoFMRthresholds,CARPM-FIQAplaces4th(CARPM-FIQA(L))and6th(CARPM-FIQA(S))of17comparedmethodsbypAUC-EDCandAUC-EDCaveragedacrossFRmodelsand,afterper-benchmarknormalization,acrossbenchmarks,stayingwithinafewpercentofthebestmethod’snormalizedaverageforeveryFRmodel,providingaprincipledsolutiontotraininginstabilitywhilemaintainingtheperformancebenefitsofFRintegration.Morebroadly,ourworkdemonstratesthattemporalaggregationstrategiescanstabilizetrainingobjectivesindeeplearningsystemswheretargetvaluesinherentlyfluctuateduetoevolvingfeaturerepresentations.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2609\.31662

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.31662 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.31662 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.31662 in a Space README.md to link it from this page.

Collections including this paper1

Similar Articles

Cautious Conflict-Free Inverse Gradient Methods

arXiv cs.LG

提出了 Cautious ConFIG，一个三行扩展的优化算法，修复了 ConFIG 与 Adam 等自适应优化器结合时破坏无冲突梯度方向的问题，并在多个 PINN 基准上取得一致性能提升。

Learning from the Gap Between Pass@K and Pass@1

arXiv cs.LG

The paper introduces GapFT, a fine-tuning method that selects training examples based on a model's Pass@K–Pass@1 gap, showing on LogiQA 2.0 and ReClor that it raises single-sample accuracy dramatically while using far less data than uniform verified RFT.