在二进制神经网络上的边缘设备的预计算批量规范化参数
Nicholas Phipps1,2, Jin-Jia Shang1,2, Tee Hui Teo1
1Engineering Product Development, Singapore University of Technology and Design, Singapore 487372, Singapore.
Sensors (Basel, Switzerland)
|July 8, 2023
概括
二元化神经网络 (BNNs) 为边缘设备优化批量规范化 (BN). 预先计算BN参数可以显著降低63%的内存使用量,而不会影响准确性.
科学领域:
- 计算机科学 计算机科学
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 二元化神经网络 (BNN) 是量子化卷积神经网络 (CNN),通过降低参数精度来减少模型大小.
- 批量规范化 (BN) 层在BNN中至关重要,但其浮点运算在边缘设备上是计算上昂贵的.
研究的目的:
- 为了减少边缘设备上的BNN的内存足迹.
- 在推理过程中优化BN层的计算效率.
主要方法:
- 在量子化之前预计算批量规范化 (BN) 参数,以便在推理过程中利用模型的固定性.
- 使用MNIST数据集实施和验证拟议的BNN方法.
主要成果:
- 减少了63%的内存使用量,实现了860字节的模型大小.
- 保持了与传统计算方法相比的准确性.
- 将BN层计算所需的循环数减少到两个边缘设备.
结论:
- 预计算BN参数是边缘设备BNN内存和计算优化的有效策略.
- 拟议的方法可以显著节省内存,而不会影响模型的准确性.
相关概念视频
01:37Neural Regulation
39.6K
Digestion begins with a cephalic phase that prepares the digestive system to receive food. When our brain processes visual or olfactory information about food, it triggers impulses in the cranial nerves innervating the salivary glands and stomach to prepare for food.
39.6K
01:25Neural Circuits
1.3K
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
1.3K