相关实验视频

Updated: Jul 26, 2025

Photodiode-Based Optical Imaging for Recording Network Dynamics with Single-Neuron Resolution in Non-Transgenic Invertebrates

Photodiode-Based Optical Imaging for Recording Network Dynamics with Single-Neuron Resolution in Non-Transgenic Invertebrates

Published on: July 9, 2020

一次性光学神经网络

Liane Bernstein¹, Alexander Sludds^1,2, Christopher Panuski¹

¹Research Laboratory of Electronics, Massachusetts Institute of Technology, 50 Vassar St, Cambridge, MA 02139, USA.

Science advances

|June 21, 2023

概括

本研究介绍了一种用于深度神经网络 (DNN) 的新型模拟处理器,该处理器利用光学进行高效的计算. 它在没有重新培训的情况下在标准数据集上实现了高精度,克服了以前的可扩展性限制.

科学领域:

光电学是指光电子产品.
人工智能的人工智能
计算机工程计算机工程

背景情况:

模拟硬件为深度神经网络 (DNN) 提供了比数字电子更有效的替代方案.
之前的模拟DNN处理器在可扩展性方面面临限制,需要重新训练模型,这阻碍了实际应用.
开发高效的硬件对于推进下一代AI和机器学习至关重要.

研究的目的:

提出一个可扩展的,CMOS兼容的模拟DNN处理器,利用自由空间光学和光电子.
用标准DNN模型和数据集来证明处理器的有效性,而无需重新培训.
通过实验确定光学带宽所造成的吞吐量限制.

主要方法:

开发了一个CMOS兼容的模拟DNN处理器,使用自由空间光学来分配输入矢量.
利用光电子来实现静态,可更新的权重和非线性.
在使用标准完全连接的DNN的MNIST,Fashion-MNIST和QuickDraw数据集上测试了处理器.

主要成果:

实现了高分类准确率:95.6% (MNIST),83.3% (时尚-MNIST) 和79.0% (快速绘制).
经过证明的每层单一射击分类,没有预处理或再培训.
通过实验确定了基本吞吐量上限约为0.9 exaMAC/s,由光学带宽限制.

结论:

开发的模拟光学DNN处理器为深度学习任务提供了可扩展和高效的解决方案.
该系统与标准DNN和数据集的兼容性促进了对AI的模拟计算的更广泛采用.
这项技术通过结合广泛的光谱和空间带宽,为高效的下一代计算架构铺平了道路.

更多相关视频

Optical Recording of Suprathreshold Neural Activity with Single-cell and Single-spike Resolution

Optical Recording of Suprathreshold Neural Activity with Single-cell and Single-spike Resolution

Published on: September 5, 2012

Single-unit In vivo Recordings from the Optic Chiasm of Rat

Single-unit In vivo Recordings from the Optic Chiasm of Rat

Published on: April 2, 2010

相关概念视频

Neural Circuits

Neural Circuits

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...

Anatomy of the Eyeball

Anatomy of the Eyeball

The eye is a spherical, hollow structure composed of three tissue layers. The outer layer — the fibrous tunic, comprises the sclera — a white structure — and the cornea, which is transparent. The sclera encompasses some of the ocular surface, most of which is not visible. However, the 'white of the eye' is distinctively visible in humans compared to other species. The cornea, a clear covering at the front of the eye, enables light penetration. The eye's middle...