# A Time Encoding approach to training Spiking Neural Networks

arXiv: https://arxiv.org/abs/2110.06735 · HTML: https://mlsanity.com/paper/2110.06735/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 27 references that are in the map.
- First posted on arXiv 13 Oct 2021.

## Builds on (most cited first, 10 of 10)

- [Surrogate Gradient Learning in Spiking Neural Networks](https://mlsanity.com/paper/1901.09948/) (2019)
- [Event-Based Backpropagation can compute Exact Gradients for Spiking Neural Networks](https://mlsanity.com/paper/2009.08378/) (2020)
- [Temporal Coding in Spiking Neural Networks with Alpha Synaptic Function: Learning with Backpropagation](https://mlsanity.com/paper/1907.13223/) (2019)
- [Learning from Event Cameras with Sparse Spiking Convolutional Neural Networks](https://mlsanity.com/paper/2104.12579/) (2021)
- [Reconstructing Classes of Non-bandlimited Signals from Time Encoded Information](https://mlsanity.com/paper/1905.03183/) (2019)
- [Sampling and Reconstruction of Bandlimited Signals with Multi-Channel Time Encoding](https://mlsanity.com/paper/1907.05673/) (2019)
- [Asynchrony Increases Efficiency: Time Encoding of Videos and Low-Rank Signals](https://mlsanity.com/paper/2104.14511/) (2021)
- [Encoding and Decoding Mixed Bandlimited Signals using Spiking Integrate-and-Fire Neurons](https://mlsanity.com/paper/1910.09413/) (2019)
- [Matrix recovery from bilinear and quadratic measurements](https://mlsanity.com/paper/2001.04933/) (2020)
- [Time Encoding of Finite-Rate-of-Innovation Signals](https://mlsanity.com/paper/2107.03344/) (2021)

## Built on by (most cited first, 1 of 1)

- [How Asynchronous Events Encode Video](https://mlsanity.com/paper/2206.04341/) (2022)

## Abstract (arXiv)

While Spiking Neural Networks (SNNs) have been gaining in popularity, it seems that the algorithms used to train them are not powerful enough to solve the same tasks as those tackled by classical Artificial Neural Networks (ANNs). In this paper, we provide an extra tool to help us understand and train SNNs by using theory from the field of time encoding. Time encoding machines (TEMs) can be used to model integrate-and-fire neurons and have well-understood reconstruction properties. We will see how one can take inspiration from the field of TEMs to interpret the spike times of SNNs as constraints on the SNNs' weight matrices. More specifically, we study how to train one-layer SNNs by solving a set of linear constraints, and how to train two-layer SNNs by leveraging the all-or-none and asynchronous properties of the spikes emitted by SNNs. These properties of spikes result in an alternative to backpropagation which is not possible in the case of simultaneous and graded activations as in classical ANNs.