基于二进制神经网络的IoT器件上的声音事件检测
Sound Event Detection with Binary Neural Networks on Tightly Power-Constrained IoT Devices
摘要:
声音事件检测(SED)是消费者和智慧城市应用中的热门话题。基于深度神经网络的现有方法非常有效,但是当针对超低功率始终开启设备时,在存储器、功率和吞吐量方面要求很高。延迟,可用性,成本和隐私要求正在推动最近的物联网系统在节点上处理数据,靠近传感器,能源供应非常有限,并且对内存大小和处理能力的严格限制排除了运行最先进的DNN。在本文中,我们探索了极端量化与小尺寸二进制神经网络(BNN)的结合,以及高能效的基于RISC-V的(8+1)核GAP 8微控制器。从SED的现有CNN开始,其占用空间(815 kB)超过了我们平台上可用的512 kB内存,我们使用二进制滤波器和激活来重新训练网络,以匹配这些内存限制。(完全)二进制神经网络在具有挑战性的ImageNet对象识别挑战中,与其等效的全精度基线相比,准确率自然下降了12-18%。这个BNN达到了77.9%的精度,仅比全精度版本低7%,权重为58 kB(少7.2倍),内存为262 kB(少2.4倍)。通过我们的BNN实现,我们在整个网络上达到了4.6 GMAC/s和1.5 GMAC/s的峰值吞吐量,包括Mel bin预处理,分别对应于67.1 GMAC/s/W和31.3 GMAC/s/W的效率。与ARM Cortex-M4实现的性能相比,我们的系统具有10.3倍的执行时间和51.1倍的能效。
英文摘要:
Sound event detection (SED) is a hot topic in consumer and smart city applications. Existing approaches based on Deep Neural Networks are very effective, but highly demanding in terms of memory, power, and throughput when targeting ultra-low power always-on devices. Latency, availability, cost, and privacy requirements are pushing recent IoT systems to process the data on the node, close to the sensor, with a very limited energy supply, and tight constraints on the memory size and processing capabilities precluding to run state-of-the-art DNNs. In this paper, we explore the combination of extreme quantization to a small-footprint binary neural network (BNN) with the highly energy-efficient, RISC-V-based (8+1)-core GAP8 microcontroller. Starting from an existing CNN for SED whose footprint (815 kB) exceeds the 512 kB of memory available on our platform, we retrain the network using binary filters and activations to match these memory constraints. (Fully) binary neural networks come with a natural drop in accuracy of 12-18% on the challenging ImageNet object recognition challenge compared to their equivalent full-precision baselines. This BNN reaches a 77.9% accuracy, just 7% lower than the full-precision version, with 58 kB (7.2 times less) for the weights and 262 kB (2.4 times less) memory in total. With our BNN implementation, we reach a peak throughput of 4.6 GMAC/s and 1.5 GMAC/s over the full network, including preprocessing with Mel bins, which corresponds to an efficiency of 67.1 GMAC/s/W and 31.3 GMAC/s/W, respectively. Compared to the performance of an ARM Cortex-M4 implementation, our system has a 10.3 times faster execution time and a 51.1 times higher energy-efficiency.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。