Practical Bayesian Inference for Speech SNNs: Uncertainty and Loss-Landscape Smoothing

Yesmine Abdennadher; Philip N. Garner

cs.LGApr 9, 2026

Practical Bayesian Inference for Speech SNNs: Uncertainty and Loss-Landscape Smoothing

Authors: Yesmine Abdennadher, Philip N. Garner

Organizations: Idiap Research Institute, Martigny, Switzerland

Abstract

Spiking Neural Networks (SNNs) are naturally suited for speech processing tasks due to their specific dynamics, which allows them to handle temporal data. However, the threshold-based generation of spikes in SNNs intuitively causes an angular or irregular predictive landscape. We explore the effect of using the Bayesian learning approach for the weights on the irregular predictive landscape. For the surrogate-gradient SNNs, we also explore the application of the Improved Variational Online Newton (IVON) approach, which is an efficient variational approach. The performance of the proposed approach is evaluated on the Heidelberg Digits and Speech Commands datasets. The hypothesis is that the Bayesian approach will result in a smoother and more regular predictive landscape, given the angular nature of the deterministic predictive landscape. The experimental evaluation of the proposed approach shows improved performance on the negative log-likelihood and Brier score. Furthermore, the proposed approach has resulted in a smoother and more regular predictive landscape compared to the deterministic approach, based on the one-dimensional slices of the weight space

Figures & tables

Figure 1. Local one-dimensional slice of the deterministic cross-entropy loss on a fixed batch.

Dataset	Method	Acc(%) $\uparrow$	NLL $\downarrow$	ECE $\downarrow$	Entropy $\downarrow$	MI $\uparrow$	Brier $\downarrow$
HD	Deterministic (MAP)	98.10	0.0553	0.0097	0.0824	–	0.0294
HD	IVON (MC $S{=}20$ )	99.51	0.0233	0.0103	0.0450	0.0163	0.0105
SC	Deterministic (MAP)	77.01	0.8063	0.0224	0.7967	–	0.3184
SC	IVON (MC $S{=}20$ )	79.93	0.7191	0.0133	0.6829	0.0711	0.2838

Table 1. Test-set results on Heidelberg Digits (HD) and Speech Commands (SC).

Figure 2. Local one-dimensional slice of the IVON Bayesian predictive loss on a fixed batch.

Explore similar work

Latency and accuracy tradeoffs in Spiking Neural Networks
Sep 28, 2026Zhanglu Yan, Zixuan Zhu, Kaiwen Tang +3Efficient Neural Network Inference Automatic Speech Recognition
Department of computer science, National University of Singapore · Shanghai Advanced Research Institute, Chinese Academy of Science · University of Chinese Academy of Sciences, Beijing +1
Linearized Bregman Iterations for Sparse Spiking Neural Networks
Mar 17, 2026Daniel Windhager, Bernhard A. Moser, Michael LunglmayrSpiking Neural Networks Sparse Neural Networks
Silicon Austria Labs, Linz, Austria · Software Competence Center Hagenberg, Hagenberg, Austria · Institute of Signal Processing, Johannes Kepler University, Linz, Austria
Adaptive Speech-to-Spike Encoding for Spiking Neural Networks
Jun 17, 2026Taharim Rahman Anon, Jakaria Islam EmonKeyword Spotting Speech Processing
1PI LLC, Sapporo, Hokkaido, Japan

Sep 28, 2026J/K move · Enter open · S save

Latency and accuracy tradeoffs in Spiking Neural Networks

Zhanglu Yan, Zixuan Zhu, Kaiwen Tang, Yuyang Cai+2

Department of computer science, National University of Singapore · Shanghai Advanced Research Institute, Chinese Academy of Science · University of Chinese Academy of Sciences, Beijing+1

Efficient Neural Network Inference Automatic Speech Recognition Spiking Neural Networks

Open paper AskPDF arXiv

Spiking neural networks are attractive for low-power speech command recognition, yet their latency has received far less attention than their energy efficiency, and their multi-timestep execution is widely assumed to make them slower than quantized neural networks. This paper challenges the assumption that more local timesteps necessarily imply higher network latency. By overlapping computation across adjacent layers at the timestep level, SNNs may complete execution in less time than comparable bit-serial QNNs. However, this overlap relies on spikes firing on incomplete inputs, and a spike once generated cannot be withdrawn, so its error persists and reduces accuracy. Waiting for more input before firing would seem to improve accuracy at the cost of reduced overlap. Yet we find and prove that this intuition fails at some layers, where even a small increase in waiting can change spike timing and downstream computation, making the network both slower and less accurate. We therefore propose a Pipeline Delay Search method which selects each layer's delay by balancing task-level accuracy gains against added network latency. We then adapt the selected configurations through spike-based quantization-aware training and bounded tuning of firing thresholds and initial membrane potentials. Together, these steps form Falcon, a framework for Fine-grained Analysis of Latency and Controlled firing which systematically analyzes and optimizes SNN latency under a spatial analog compute-in-memory mapping with shared digital engines. We evaluate Falcon on GSCV2 and SSC, achieving competitive accuracies of 96.31 and 83.02 at modeled network-core latencies of 119.64 and 124.00us, respectively. Together, our analysis and results show that SNNs can compute more yet finish faster, and wait longer yet predict worse, highlighting why Falcon matters for both latency and accuracy.

Abstract

Explore similar work

Latency and accuracy tradeoffs in Spiking Neural Networks

Linearized Bregman Iterations for Sparse Spiking Neural Networks

Adaptive Speech-to-Spike Encoding for Spiking Neural Networks