What Does 4-Bit Microscaling Cost a Wearable Activity Recognizer?
作者: Megan Healy, Hossein Valavi
分类: cs.LG
发布日期: 2026-11-03
💡 一句话要点
评估4位微缩对可穿戴活动识别器的影响
🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)
关键词: 可穿戴设备 活动识别 低精度计算 深度学习 模型优化 4位微缩 运动传感器
📋 核心要点
- 现有的可穿戴传感器模型在精度与存储效率之间存在权衡，尤其是在低精度数值表示时。
- 论文提出了使用MXFP4和NVFP4两种4位块缩放格式来优化可穿戴活动识别器的性能。
- 实验结果表明，NVFP4在多个测试折中均优于MXFP4，且通过调整缩放规则可显著提升模型性能。
📝 摘要（中文）
可穿戴运动传感器能够在设备上识别动作，降低精度的数值可以减少模型存储和内存流量。本文评估了两种4位块缩放格式MXFP4和NVFP4对小型可穿戴传感器模型的影响。通过训练一个159,492参数的1D CNN模型识别四种篮球动作，结果显示在全精度下模型达到了69.26%的宏F1值，而4位权重的使用导致了1.45到5.14个百分点的性能损失。NVFP4在所有测试中表现优于MXFP4，且通过调整缩放规则可以显著改善性能。
🔬 方法详解
问题定义：本文旨在评估4位微缩格式对小型可穿戴活动识别模型的影响，现有方法在低精度表示下可能导致性能损失。
核心思路：通过比较MXFP4和NVFP4两种4位块缩放格式，探索其对模型性能的影响，特别是在可穿戴传感器应用中的有效性。
技术框架：研究中训练了一个159,492参数的1D CNN模型，使用五折交叉验证对四种篮球动作进行识别，比较不同数值格式的性能。
关键创新：本文首次测量了4位微缩格式在可穿戴传感器模型中的具体成本，揭示了MXFP4和NVFP4在性能上的显著差异。
关键设计：实验中采用了不同的块大小和缩放规则，发现MXFP4的性能损失主要源于缩放选择不当，调整缩放规则后可显著改善模型表现。使用8位浮点缩放进一步提升了性能。
📊 实验亮点
实验结果显示，在全精度下模型达到了69.26%的宏F1值，而使用4位权重的MXFP4和NVFP4分别损失了1.45和0.72个百分点。NVFP4在所有五个测试折中均表现优于MXFP4，且通过优化缩放规则，性能提升可达3.63个百分点。
🎯 应用场景
该研究的潜在应用领域包括智能手环、运动监测设备及其他可穿戴技术，能够在保持较低存储需求的同时提高活动识别的准确性，具有重要的实际价值和未来影响。
📄 摘要（原文）
Wearable sports sensors can recognize actions on the device itself, where lower-precision numbers reduce model storage and memory traffic. Two 4-bit block-scaled formats, MXFP4 from the Open Compute Project, introduced in 2023, and NVFP4 from NVIDIA, introduced in 2025, are drawing interest. Published evaluations cover general deep-learning benchmarks and large language models; to our knowledge, nobody has yet measured what they cost a small wearable-sensor model. We train a 159,492-parameter 1D CNN to recognize four basketball actions from wrist-accelerometer data in Hang-Time HAR and test it on players it has never seen, using five-fold cross-validation grouped by player. Every comparison between number formats is made fold by fold. In full precision the model reaches 69.26\% macro-F1. Measured TFLite FP16 and INT8 models lose at most 0.16 percentage points. In simulation, 4-bit weights cost 1.45 points with MXFP4 and 0.72 points with NVFP4. When activations are quantized to 4 bits as well, MXFP4 loses 5.14 points and NVFP4 only 0.75, and NVFP4 comes out ahead in all five held-out-player folds. The ranking survives two alternative simulator settings, with NVFP4 leading by 3.49 to 4.40 points. A $2 \times 3$ ablation with 4-bit weights and activations (two block sizes and three scale rules) points to scale selection as the main source of MXFP4's loss. The reference MXFP4 conversion we simulate rounds the scale down, which can pick scales that are too small, so many blocks clip their largest values. Rounding the scale up recovers 3.18 points with 32-value blocks and 3.63 points with 16-value blocks, and switching to an 8-bit floating-point scale, as NVFP4 does, adds another 0.84 and 1.20 points. Block size matters less. Under this round-down rule, 16-value blocks score 0.44 points below 32-value blocks in all five folds; under the other two rules there is no consistent effect.