解决模型在多领域和长尾数据下的量化难题，提升边缘设备推理效率。
Toward Multi-Domain and Long-Tailed Quantization via Feature Alignment and Scaling
- 通过累积分布投影对齐不同领域特征分布，稳定多域量化过程。
- 在低比特下仍保持高精度，长尾数据上性能优于现有方法。
- 适合部署于复杂真实场景的轻量级模型，如移动设备与物联网。
深度神经网络量化对资源受限设备的高效推理至关重要。然而，现有方法多针对单一领域与类别平衡的数据，对实际中常见的领域偏移或严重类别不平衡问题关注不足。本文提出高效多领域对齐量化（EmaQ），通过基于累积分布函数的投影对齐各领域特征分布，并采用敏感度感知的权重聚合策略以稳定多域量化。进一步扩展至长尾量化方法EmaQ-LT，引入类别条件方差缩放与置信度感知的logit调整，缓解多数类过自信问题。理论分析提供了收敛保证，并支撑所提机制的设计。在标准数据集及多领域（Office-31、Digits）和长尾数据集（SynDigits-LT、CIFAR-10-LT、CIFAR-100-LT）上的实验表明，EmaQ与EmaQ-LT在领域迁移和类别不平衡条件下均实现优异的低比特性能。
原文摘要 · Abstract (English)
Quantizing deep neural networks is essential for efficient inference on resource-constrained devices. However, most existing methods are designed for single-domain and class-balanced data, leaving practical settings with domain shifts or severe class imbalance underexplored. We address these challenges with Efficient Multi-Domain Alignment Quantization (EmaQ), which aligns domain distributions through a CDF-based projection and uses sensitivity-aware weight aggregation to stabilize multi-domain quantization. We further extend EmaQ to EmaQ-LT for long-tailed quantization by introducing class-conditioned variance scaling and confidence-based logit adjustment to mitigate majority-class overconfidence. Theoretical analyses establish convergence guarantees and motivate the proposed sensitivity and scaling mechanisms. Experiments on standard, multi-domain (Office-31, Digits), and long-tailed (SynDigits-LT, CIFAR-10-LT, CIFAR-100-LT) benchmarks show that EmaQ and EmaQ-LT achieve strong low-bit performance under domain shift and class imbalance.
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。