解决行人属性识别中稀有属性难学习问题，提出轻量级损失优化方案。
An Empirical Analysis of Optimization Dynamics and Sparsity Boundaries in Large-Scale Pedestrian Attribute Recognition
- 用改进的Focal Loss缓解少数类样本过少导致的模型偏差。
- 在0.1%以下正样本率时，传统重加权失效，需逐样本干预。
- 仅调整损失函数，无额外计算开销，适合边缘设备部署。
行人属性识别（PAR）对视频监控中的罪案检索和重识别至关重要。将PETA与PA-100K合并成包含109,000张图像的综合数据集时，极端类别不平衡问题依然存在，少数属性正样本比例低于1%，导致标准二值交叉熵（BCE）优化会压制罕见属性，我们称之为‘多数负类欺骗陷阱’。本文系统分析了多标签Focal Loss超参数（alpha与gamma）在ResNet-18上的表现，发现校准配置（alpha=0.50, gamma=2.0）可实现62.32%的宏平均F1分数，与BCE基线相当，同时具备更优的困难样本挖掘能力和收敛性。该方法仅通过损失函数工程实现，无额外计算开销，适用于边缘部署。我们还揭示了‘稀疏墙’现象：当正样本比例低于0.1%时，全局损失重加权失效，必须采用实例级干预策略。
原文摘要 · Abstract (English)
Pedestrian Attribute Recognition (PAR) is critical for video surveillance, enabling forensic search and re-identification systems. Extreme class imbalance remains a fundamental obstacle when merging PETA and PA-100K into a 109,000-image composite corpus, where minority attributes have positive sample fractions below 1%. This causes standard BCE optimization to suppress rare traits, a phenomenon we term the majority negative class cheating trap. We present a systematic ablation of Multi-Label Focal Loss hyperparameters (alpha and gamma) on a ResNet-18 backbone. A calibrated configuration (alpha=0.50, gamma=2.0) achieves a Macro F1-score of 62.32%, matching BCE baseline while preserving superior hard-example mining and convergence dynamics. Our approach uses pure loss-function engineering with zero computational overhead for edge deployment. We identify the Sparsity Wall, a hard boundary where positive sample fractions below 0.1% make global loss reweighting ineffective, requiring instance-level intervention.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。