4位量化下仍可达到12位原始数据效果，实现低功耗模式识别。
$γ$-Quant: Towards Learnable Quantization for Low-bit Pattern Recognition
- 针对低比特传感场景，提出可学习的非线性γ量化方法
- 4位量化在目标检测与可穿戴设备动作识别上媲美12位原始数据
- 适合资源受限设备的实时模式识别应用
多数模式识别模型基于预处理数据开发。在计算机视觉中，图像分析网络通常接收经过图像信号处理（ISP）流水线处理的RGB图像，该流程旨在优化人眼感知。然而，许多现代视觉任务无需人工参与，这引发疑问：此类预处理是否仍对自动化分析最优？类似地，可穿戴传感器上的动作识别通常采用高比特模数转换器（ADC）生成的归一化浮点数据作为输入，但这种方式在数据传输效率上极不经济，显著影响可穿戴设备电池寿命。本文聚焦于低带宽与低功耗场景，即传感器仅支持低比特深度采集。我们提出γ-Quant——一种面向特定任务的可学习非线性量化方法。以原始图像目标检测和可穿戴设备动作识别为例，验证了仅用4位量化即可达到与12位原始数据相当的性能。所有实验代码已公开于https://github.com/Mishalfatima/Gamma-Quant。
原文摘要 · Abstract (English)
Most pattern recognition models are developed on pre-proce\-ssed data. In computer vision, for instance, RGB images processed through image signal processing (ISP) pipelines designed to cater to human perception are the most frequent input to image analysis networks. However, many modern vision tasks operate without a human in the loop, raising the question of whether such pre-processing is optimal for automated analysis. Similarly, human activity recognition (HAR) on body-worn sensor data commonly takes normalized floating-point data arising from a high-bit analog-to-digital converter (ADC) as an input, despite such an approach being highly inefficient in terms of data transmission, significantly affecting the battery life of wearable devices. In this work, we target low-bandwidth and energy-constrained settings where sensors are limited to low-bit-depth capture. We propose $γ$-Quant, i.e.~the task-specific learning of a non-linear quantization for pattern recognition. We exemplify our approach on raw-image object detection as well as HAR of wearable data, and demonstrate that raw data with a learnable quantization using as few as 4-bits can perform on par with the use of raw 12-bit data. All code to reproduce our experiments is publicly available via https://github.com/Mishalfatima/Gamma-Quant
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。