David Chesmore

Automated bioacoustic identification of species

Research into the automated identification of animals by bioacoustics is becoming more widespread mainly due to difficulties in carrying out manual surveys. This paper describes automated recognition of insects (Orthoptera) using time domain signal coding and artificial neural networks. Results of field recordings made in the UK in 2002 are presented which show that it is possible to accurately recognize 4 British Orthoptera species in natural conditions under high levels of interference. Work is under way to increase the number of species recognized.

Development of non‐invasive monitoring methods for larvae and adults of the stag beetle, Lucanus cervus

Abstract.  1. The stag beetle, Lucanus cervus is Nationally Scarce in the UK, yet no methods exist for monitoring the abundance of adults or presence of the subterranean larvae.

Bioinspired Load Balancing in Large-Scale WSNs Using Pheromone Signalling

Wireless sensor networks (WSNs) consist of multiple, distributed nodes each with limited resources. With their strict resource constraints and application-specific characteristics, WSNs contain many challenging tradeoffs. This paper proposes a bioinspired load balancing approach, based on pheromone signalling mechanisms, to solve the tradeoff between service availability and energy consumption. We explore the performance consequences of the pheromone-based load balancing approach using (1) a system-level simulator, (2) deployment of real sensor testbeds to provide a competitive analysis of these evaluation methodologies. The effectiveness of the proposed algorithm is evaluated with different scenario parameters and the required performance evaluation techniques are investigated on case studies based on sound sensors.

Runtime optimisation in WSNs for load balancing using pheromone signalling

Wireless Sensor Networks (WSNs) consist of multiple, distributed nodes each with limited resources. With their strict resource constraints and application-specific characteristics, WSNs contain many challenging trade-offs. This paper proposes a bio-inspired load balancing approach, based on pheromone signalling mechanisms, to solve the trade-off between service availability and energy consumption. We explore the performance consequences of the pheromone-based load balancing approach using: 1) a system-level simulator; 2) deployment of real sensor testbeds to provide a competitive analysis of these evaluation methodologies. The effectiveness of the proposed algorithm is evaluated with different scenario parameters and the required performance evaluation techniques are investigated on case studies based on sound sensors.

Masked Conditional Neural Networks for Audio Classification

We present the ConditionaL Neural Network (CLNN) and the Masked ConditionaL Neural Network (MCLNN) designed for temporal signal recognition. The CLNN takes into consideration the temporal nature of the sound signal and the MCLNN extends upon the CLNN through a binary mask to preserve the spatial locality of the features and allows an automated exploration of the features combination analogous to hand-crafting the most relevant features for the recognition task. MCLNN has achieved competitive recognition accuracies on the GTZAN and the ISMIR2004 music datasets that surpass several state-of-the-art neural network based architectures and hand-crafted methods applied on both datasets.

Automated acoustic identification of beetle larvae in imported goods using time domain analysis

The detection of insect pests in imported goods is of considerable economic importance and the automation of this process is becoming more viable both technologically and financially. As a result, the Department for Environment, Food and Rural Affairs in the UK has funded a research project to develop instrumentation facilitating real‐time acoustic detection of the feeding activity of insect larvae inside imported goods, such as timber. The instrumentation will also be capable of species‐level identification. Previous work at York has shown that detection of beetle larvae in wood is possible using low cost piezoelectric sensors. The project described here extends this work by investigating a number of signal analysis methods for robust detection of biting events, including fractal dimension analysis. Identification is currently being carried out using time domain signal coding and artificial neural networks. This paper will concentrate on the results of various algorithms for the estimation of fractal dimension and their relative suitability for bite detection. The effects of varying sampling rates, threshold levels and signal‐to‐noise ratio on the detection rate will be demonstrated.

Masked Conditional Neural Networks for Environmental Sound Classification

The ConditionaL Neural Network (CLNN) exploits the nature of the temporal sequencing of the sound signal represented in a spectrogram, and its variant the Masked ConditionaL Neural Network (MCLNN) induces the network to learn in frequency bands by embedding a filterbank-like sparseness over the networks links using a binary mask. Additionally, the masking automates the exploration of different feature combinations concurrently analogous to handcrafting the optimum combination of features for a recognition task. We have evaluated the MCLNN performance using the Urbansound8k dataset of environmental sounds. Additionally, we present a collection of manually recorded sounds for rail and road traffic, YorNoise, to investigate the confusion rates among machine generated sounds possessing low-frequency components. MCLNN has achieved competitive results without augmentation and using 12% of the trainable parameters utilized by an equivalent model based on state-of-the-art Convolutional Neural Networks on the Urbansound8k. We extended the Urbansound8k dataset with YorNoise, where experiments have shown that common tonal properties affect the classification performance.

Automated Identification of Vehicles using Acoustic Signal Processing

Automated identification of unauthorised intruding vehicles approaching protected infrastructure is becoming increasingly important for security purposes. This three-year project, which is now in its second year, aims to develop a real-time acoustic vehicle type recognition system that will be predominantly composed of three main parts; acoustic signal pre-processing, feature extraction, and decision making (or classification). The main study area covers various signal processing techniques in time, time-frequency and potentially frequency domains with signal classification implemented using a range of artificial intelligence techniques such as artificial neural networks. So far the focus has been on time domain signal processing and neural network classification. Whilst the work is at an early stage, the time domain methods such as Time Domain Signal Coding (TDSC) and Co-Occurrence Matrix combined with neural networks have already shown some promise. The presentation will introduce the project, describe methodologies involved as well as the results to date for the novel acoustic identification of categories of vehicles.

Music Genre Classification Using Masked Conditional Neural Networks

The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The CLNN captures the conditional temporal influence between the frames in a window and the mask in the MCLNN enforces a systematic sparseness that follows a filterbank-like pattern over the network links. The mask induces the network to learn about time-frequency representations in bands, allowing the network to sustain frequency shifts. Additionally, the mask in the MCLNN automates the exploration of a range of feature combinations, usually done through an exhaustive manual search. We have evaluated the MCLNN performance using the Ballroom and Homburg datasets of music genres. MCLNN has achieved accuracies that are competitive to state-of-the-art handcrafted attempts in addition to models based on Convolutional Neural Networks.

Environmental Sound Recognition Using Masked Conditional Neural Networks.

Neural network based architectures used for sound recognition are usually adapted from other application domains, which may not harness sound related properties. The ConditionaL Neural Network (CLNN) is designed to consider the relational properties across frames in a temporal signal, and its extension the Masked ConditionaL Neural Network (MCLNN) embeds a filterbank behavior within the network, which enforces the network to learn in frequency bands rather than bins. Additionally, it automates the exploration of different feature combinations analogous to handcrafting the optimum combination of features for a recognition task. We applied the MCLNN to the environmental sounds of the ESC-10 dataset. The MCLNN achieved competitive accuracies compared to state-of-the-art convolutional neural networks and hand-crafted attempts.