用人脸数量衡量数据难度，发现密度越高模型越难准确计数。
Face Density as a Proxy for Data Complexity: Quantifying the Hardness of Instance Count
- 通过控制类别不平衡，只改变每图人脸数来量化密度影响。
- 人脸数量从1到18时，模型错误率最高上升4.6倍，性能持续下降。
- 低密度训练的模型无法适应高密度场景，提示需分密度评估与训练。
机器学习进展长期聚焦于模型改进，但实际性能常受限于数据本身的复杂性。本文将实例密度（以人脸数量衡量）作为数据复杂性的核心驱动因素进行分离与量化。不同于简单观察‘人群场景更难’，我们严格控制类别不平衡，在WIDER FACE和Open Images数据集上仅保留每张图像1至18个脸且采样完全平衡，结果表明模型性能随人脸数量增加而单调下降。这一趋势在分类、回归和检测任务中均成立，即使模型全程接触全部密度范围亦然。此外，仅在低密度下训练的模型在高密度场景下泛化失败，表现出系统性低估偏差，错误率最高达4.6倍，表明密度变化构成域偏移。研究确立了实例密度为数据难度的内在可量化维度，推动课程学习与分密度评估等针对性干预。
原文摘要 · Abstract (English)
Machine learning progress has historically prioritized model-centric innovations, yet achievable performance is frequently capped by the intrinsic complexity of the data itself. In this work, we isolate and quantify the impact of instance density (measured by face count) as a primary driver of data complexity. Rather than simply observing that ``crowded scenes are harder,'' we rigorously control for class imbalance to measure the precise degradation caused by density alone. Controlled experiments on the WIDER FACE and Open Images datasets, restricted to exactly 1 to 18 faces per image with perfectly balanced sampling, reveal that model performance degrades monotonically with increasing face count. This trend holds across classification, regression, and detection paradigms, even when models are fully exposed to the entire density range. Furthermore, we demonstrate that models trained on low-density regimes fail to generalize to higher densities, exhibiting a systematic under-counting bias, with error rates increasing by up to 4.6x, which suggests density acts as a domain shift. These findings establish instance density as an intrinsic, quantifiable dimension of data hardness and motivate specific interventions in curriculum learning and density-stratified evaluation.
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。