Unlocking spiking neural networks for machine learning research

解锁用于机器学习研究的尖峰神经网络

• 批准号: EP/V052241/1
• 负责人: James Knight
• 金额: $ 106.36万
• 依托单位: University of Sussex
• 依托单位国家: 英国
• 项目类别: Fellowship
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=EP%2FV052241%2F1
• 关键词: Unlocking; spiking; neural; networks; machine

In the last decade there has been an explosion in artificial intelligence research in which artificial neural networks, emulating biological brains, are used to solve problems ranging from obstacle avoidance in self-driving cars to playing complex strategy games. This has been driven by mathematical advances and powerful new computer hardware which has allowed large 'deep networks' to be trained on huge amounts of data. For example, after training a deep network on 'ImageNet' - which consists of over 14 million manually annotated images - it can accurately identify the content of images. However, while these deep networks have been shown to learn similar patterns of connections to those found in the parts of our brains responsible for early visual processing, they differ from real brains in several important ways, especially in how the individual neurons communicate. Neurons in real brains exchange information using relatively infrequent electrical pulses known as 'spikes', whereas, in typical artificial neural network models, the spikes are abstracted away and values representing the 'rates' at which spikes would be emitted are continuously exchanged instead. However, neuroscientists believe that large amounts of information is transmitted in the precise times at which spikes are produced. Artificial 'spiking neural networks' can harness these properties, making them useful in applications which are challenging for current models such as real-world robotics and processing data with a temporal component, such as video. However, spiking neural networks can only be used effectively if suitable computer hardware and software is available. While there is existing software for simulating spiking neural networks, it has mostly been designed for studying real brains, rather than building AI systems. In this project, I am going to build a new software package which bridges this gap. It will use abstractions and processes familiar to machine learning researchers, but with techniques developed for brain simulation, allowing exciting new SNN models to be used by AI researchers. We will also explore how spiking models can be used with a special new type of sensors which directly outputs spikes rather than a stream of images. In the first phase of the project, I will focus on using Graphics Processing Units to accelerate spiking neuron networks. These devices were originally developed to speed up 3D games but have evolved into general purpose devices, widely used to accelerate scientific and AI applications. However, while these devices have become incredibly powerful and are well-suited to processing lots of data simultaneously, they are less suited to 'live' applications such as when video must be processed as fast as possible. In these situations, Field Programmable Gate Arrays - devices where the hardware itself can be re-programmed - can be significantly faster and are already being used behind the scenes in data centres. In this project, by incorporating support for FPGAs into our new software, we will make these devices more accessible to AI researchers and unlock new possibilities of using biologically-inspired spiking neural networks to learn in real-time.As well as working on these new research strands, I will also dedicate time during my fellowship to advocate for research software engineering as a valuable component of academic institutions, both via knowledge exchange and research funding. In the shorter term, I will work to develop a community of researchers involved in writing software at Sussex by organising an informal monthly 'surgery' as well as delivering specialised training on programming Graphics Processing Units and more fundamental computational and programming training for new PhD students. Finally, I will develop internship and career development opportunities for undergraduate students, to gain experience in research software engineering.

在过去的十年里，人工智能研究出现了爆炸式增长，其中模仿生物大脑的人工神经网络被用来解决从自动驾驶汽车避障到玩复杂策略游戏等问题。这是由数学进步和强大的新计算机硬件驱动的，这些硬件允许大型“深度网络”在大量数据上进行训练。例如，在“ImageNet”上训练深度网络（由超过1400万张手动注释的图像组成）后，它可以准确地识别图像的内容。然而，尽管这些深层网络已经被证明可以学习与我们大脑中负责早期视觉处理的部分相似的连接模式，但它们在几个重要方面与真实的大脑不同，特别是在单个神经元如何交流方面。真实的大脑中的神经元使用相对不频繁的被称为“尖峰”的电脉冲来交换信息，而在典型的人工神经网络模型中，尖峰被抽象掉，并且代表尖峰将被发射的“速率”的值被连续地交换。然而，神经科学家认为，大量信息是在尖峰信号产生的精确时间传输的。人工“尖峰神经网络”可以利用这些特性，使它们在对当前模型具有挑战性的应用中非常有用，例如现实世界的机器人技术和处理具有时间分量的数据，例如视频。然而，脉冲神经网络只能有效地使用，如果合适的计算机硬件和软件是可用的。虽然现有的软件用于模拟尖峰神经网络，但它主要是为了研究真实的大脑而设计的，而不是构建人工智能系统。在这个项目中，我将建立一个新的软件包，弥合这一差距。它将使用机器学习研究人员熟悉的抽象和过程，但使用为大脑模拟开发的技术，允许人工智能研究人员使用令人兴奋的新SNN模型。我们还将探讨如何使用一种特殊的新型传感器，直接输出尖峰，而不是图像流的尖峰模型。在项目的第一阶段，我将专注于使用图形处理单元来加速尖峰神经元网络。这些设备最初是为了加速3D游戏而开发的，但现在已经发展成为通用设备，广泛用于加速科学和人工智能应用。然而，虽然这些设备已经变得非常强大，并且非常适合同时处理大量数据，但它们不太适合“实时”应用，例如必须尽快处理视频的应用。在这些情况下，现场可编程门阵列-硬件本身可以重新编程的设备-可以显着更快，并且已经在数据中心的幕后使用。在这个项目中，通过将对FPGA的支持纳入我们的新软件，我们将使这些设备更容易被人工智能研究人员使用，并解锁使用生物启发的尖峰神经网络进行实时学习的新可能性。除了致力于这些新的研究领域，我还将在我的奖学金期间投入时间倡导研究软件工程作为学术机构的一个有价值的组成部分，通过知识交流和研究资助。在短期内，我将致力于开发一个社区的研究人员参与编写软件在苏塞克斯通过组织一个非正式的每月“手术”，以及提供专门的培训编程图形处理单元和更基本的计算和编程培训的新博士生。最后，我将为本科生开发实习和职业发展机会，以获得研究软件工程的经验。

项目成果

Easy and efficient spike-based Machine Learning with mlGeNN

• DOI: 10.1145/3584954.3585001
• 发表时间: 2023-04
• 期刊: Proceedings of the 2023 Annual Neuro-Inspired Computational Elements Conference
• 作者: James C. Knight;Thomas Nowotny

Efficient GPU training of LSNNs using eProp

使用 eProp 对 LSNN 进行高效 GPU 训练

• DOI: 10.1145/3517343.3517346
• 发表时间: 2022
• 作者: Knight J

Towards Autonomous Robotic Systems - 24th Annual Conference, TAROS 2023, Cambridge, UK, September 13-15, 2023, Proceedings

迈向自主机器人系统 - 第 24 届年会，TAROS 2023，英国剑桥，2023 年 9 月 13-15 日，会议记录

• DOI: 10.1007/978-3-031-43360-3_12
• 发表时间: 2023
• 作者: West A

Neural responses to reconstructed target pursuits

对重建目标追求的神经反应

• DOI: 10.1101/2023.05.03.539331
• 发表时间: 2023
• 作者: Ogawa Y

Insect-inspired Spatio-temporal Downsampling of Event-based Input

基于事件输入的受昆虫启发的时空下采样

• DOI: 10.1145/3589737.3605994
• 发表时间: 2023
• 作者: Ghosh A