Despite the advancements in cutting-edge technologies, audio signal processing continues to pose challenges and lacks the precision of a human speech processing system. To address these challenges, we propose a novel approach to simplify audio signal processing by leveraging time-domain techniques and reservoir computing. Through our research, we have developed a real-time audio signal processing system by simplifying audio signal processing through the utilization of reservoir computers, which are significantly easier to train. Feature extraction is a fundamental step in speech signal processing, with Mel Frequency Cepstral Coefficients (MFCCs) being a dominant choice due to their perceptual relevance to human hearing. However, conventional MFCC extraction relies on computationally intensive time-frequency transformations, limiting efficiency in real-time applications. To address this, we propose a novel approach that leverages reservoir computing to streamline MFCC extraction. By replacing traditional frequency-domain conversions with convolution operations, we eliminate the need for complex transformations while maintaining feature discriminability. We present an end-to-end audio processing framework that integrates this method, demonstrating its potential for efficient and real-time speech analysis. Our results contribute to the advancement of energy-efficient audio processing technologies, enabling seamless deployment in embedded systems and voice-driven applications. This work bridges the gap between biologically inspired feature extraction and modern neuromorphic computing, offering a scalable solution for next-generation speech recognition systems.
Figures & tables
Fig. 1: Topology of Reservoir computer
Fig. 2: MFC extraction
Fig. 3: Mel filter bank
Fig. 4: A comparison between the conventional approach (a) where the reservoir is used as a classifier after complex pre-processing by different means and the approach proposed here (b) where RC is used as an end-to-end audio processing concept.
Fig. 5: Time-domain filterbank
parameter=
0.002454697
0.004909393
0.00736409
0.009818787
Δf =
131
141
151
161
parameter=
0.007781114
0.00561907
0.003457026
0.001294982
Δf =
171
181
191
201
TABLE I: Parameters and frequencies for 10 Mel frequencies used
Fig. 6: Time-domain MFCC extraction
Experiment
Reservoir
Training and Testing Function
Experiment 1 (Exp1)
2
Matlab MFCC
Experiment 2 (Exp2)
1, 2
Time domain MFCC
TABLE II: Experiments
Models
Accuracy (%)
LSM [ 24 ]
94.0
Liquid-SNN [ 20 ]
77.7
Reservoir Computing (MEMS) [ 6 ]
78
Reservoir-based(Our method(EXp 1))
92.9
Reservoir-based(Our method(EXp 2))
91.82
TABLE III: Comparison of Performance of models with Ti-46 dataset for digit recognition
Models
Accuracy (%)
CNN [ 19 ]
96.4
LSTM [ 19 ]
95.23
AudioNet(Deep-NN) [ 4 ]
92.53
Liquid-SNN [ 20 ]
82.65
Reservoir-based(Our method(EXp 1))
93.08
Reservoir-based(Our method(EXp 2))
84.81
TABLE IV: Comparison of Performance of models with Audio-Mnist dataset for digit recognition.
Fig. 7: Speaker Recognition performance of our system
Fig. 8: Digit Recognition performance of our system
This paper evaluates Reservoir Computing (RC) as an autonomous, 'feature-free' framework for audio processing, designed to eliminate traditional, handcrafted feature extraction stages. We investigate whether the high-dimensional temporal dynamics inherent in a reservoir can function as a robust end-to-end processor for the direct classification of raw acoustic signals. By bypassing computationally intensive representations like MFCCs, this approach seeks to mitigate significant intellectual and pre-processing bottlenecks in traditional signal pipelines. Our study evaluates and compares shallow, sequential, and parallel deep reservoir architectures to determine their capacity for hierarchical feature representation. Experimental results demonstrate that the proposed parallel approach consistently outperforms shallow and sequential baselines while maintaining low model complexity. These findings highlight the potential of RC as an efficient and scalable alternative for time-domain audio processing, offering a promising pathway toward deployable, low-power acoustic systems with minimal preprocessing requirements.
Efficient processing of continuous audio streams remains a key challenge for real-time and resource-constrained systems. This paper introduces a neuromorphic trigger for audio event detection, based on a spiking neural network (SNN) that selectively gates input to downstream models. The proposed neuromorphic trigger acts as a flexible low-cost front-end, identifying salient audio segments and enabling these to be processed by a more computationally intensive model for tasks such as classification. The trigger is implemented as a lightweight fully connected SNN using a close-open filter for postprocessing, and is evaluated on two representative tasks: Anomalous Sound Detection (ASD) and Sound Event Detection (SED). For ASD, the trigger achieves a one-second segment-based F1 score of 0.97 on a class-agnostic form of the URBAN-SED dataset, demonstrating high reliability in identifying relevant audio regions. For SED, the trigger is combined with the Dang classifier on the DCASE 2017 Challenge Task 2 dataset, showing a potential 42.6× reduction in FLOPs while reducing the lower bound of the event-based error rate from 0.41 to 0.25. These results highlight the potential of neuromorphic triggers as real-time, energy-efficient front-end filters, enabling substantial reductions in computational cost.
In this paper, we investigate untrained recurrent models from the Reservoir Computing (RC) paradigm for audio surveillance, focusing on bidirectional Echo State Networks with different depths, from shallow to deep configurations, for emergency sound event detection. We evaluate these models on the MIVIA Audio Events dataset in a multiclass setting across different Signal-to-Noise Ratio (SNR) levels, with the goal of assessing the trade-off between depth, recognition performance, and computational efficiency. We compare the proposed architectures against fully trained recurrent and convolutional-recurrent baselines, namely Bidirectional Long Short-Term Memory networks (BiLSTMs) and Convolutional Recurrent Neural Networks (CRNNs). Results show that deep and shallow reservoir-based models achieve competitive recognition rates, with deeper variants being more robust in highly noisy conditions and shallower ones offering the most favorable efficiency profile, particularly on edge devices such as the NVIDIA Orin. In addition, the proposed approach remains robust across different input representations, including log-Mel spectrograms and MFCCs with varying resolutions. These findings highlight untrained reservoir architectures as a promising solution for resource-constrained audio surveillance scenarios.
Corrado Baccheschi, Patrizio Dazzi
Department of Computer Science · University of Pisa · Pisa, Italy