A key component in the processing of speech is the division of longer input sounds into a number of smaller sections. For speech interpretation it is generally easier to classify single sections. Similarly, when processing speech for other purposes (e.g. speech filtering), it can be easier and more relevant to process individual phonemes. Here, we propose a biologically inspired speech segmentation technique that filters the speech into multiple bandpassed channels using a Gammatone filterbank, and then uses an essentially energy-based spike coding technique in order to find the onsets and offsets present in an audio signal. These onsets and offsets are then processed using leaky integrate-and-fire neurons, and the spikes from these used to determine the speech segmentation. We evaluate this new system using a quantitative evaluation metric, and the promising results of segmentation of both clean speech and speech in noise demonstrate the effectiveness of this technique.

This paper presents the design of a CMOS cochlea filter channel which achieves high tunability, sharp stopband cutoff and low power consumption with the use of floating active inductor (FAI) as the basic building block. Simulation results show that over 40dB of gain enhancement together with 20% frequency tuning can be achieved at the same time by adjusting only one circuit parameter. A fifth-order elliptic filter providing a stop-band slope of 65.4 ~ 139.8 dB/octave is used as the last stage of the cochlea filter. The power consumption of the cochlea filter channel is 86μW.

This paper proposes a low-noise MEMS interface circuit which has very small parasitic capacitance at the input node. The circuit presented is suitable for the MEMS cochlea-mimicking acoustic sensors which are highly parasitic-sensitive due to their low intrinsic sensing capacitance. In order to reduce the electronic noise of the interface circuit, chopper stabilization technique is implemented, and an effective method to optimize the critical transistor size for best noise performance is derived. Simulation results show that, for a MEMS sensing structure with 200 fF static capacitance, the interface circuit achieves a 0.72 aF equivalent capacitance noise floor over 100 Hz to 20 kHz audio bandwidth.

A biologically-inspired neural coding scheme for the early auditory system is outlined. The cochlea response is simulated with a passive gammatone filterbank. The output of each bandpass filter is spike-encoded using a zero-crossing based method over a range of sensitivity levels. The scheme is inspired by the highly parallellised nature of the auditory nerve innervation within the cochlea. A key aspect of early auditory processing is simulated, namely that of onset detection, using leaky integrate-and-fire neuron models. Finally, a time-domain neural network (the echo state network) is used to tackle the what task of auditory perception using the output of the onset detection neurons alone. A set of interim results are presented.

This paper presents the design of a spike event coded resonant gate transistor microphone system for neuromorphic auditory applications. The microphone system employs an array of resonant gate transistors (RGT) to transduce acoustic input directly into bandpass filtered analog outputs. The bandpass filtered analog outputs are encoded as spike time events by a spike event coder and are then transmitted asynchronously by using the Address Event Representation (AER) protocol. The microphone system is designed to receive external inputs in the spike time domain to actively control the RGT response, a feature not present in other MEMS microphone systems implemented so far. System level simulations showing the response of the RGT sensor model and its spike event coded response are presented.

To localize a seen object, the superior colliculus of the barn owl integrates the visual and auditory localization cues which are accessed from the sensory system of the brain. These cues are formed as visual and auditory maps, thus the alignment between visual and auditory maps is very important for accurate localization in prey behavior. Blindness or prism wearing may disturb this alignment. The juvenile barn owl could adapt its auditory map to this mismatch after several weeks training. Here we investigate this process by building a computational model of auditory and visual integration with map adjustment in the deep superior colliculus. The adaptation is based on activity dependent axon developing which is instructed by an inhibitory network. In the inhibitory network, the strength of the inhibition is adjusted by spike timing dependent plasticity(STDP). The simulation results are in line with the biological experiment and support the idea that the STDP is involved in the alignment of sensory maps. The system of the model provides a new mechanism capable of eliminating the disparity in visual and auditory map integration.

Using a mockup of a flat panel display, with two omnidirectional microphones we have used intermicrophone time difference (ITD) to determine the azimuthal direction of a sound source (speaker). The speaker is in a noise field in an office-type environment. Bandpass filtering followed by a biologically inspired low latency onset detector (which can cope with a considerable dynamic range) determines the intervals and spectral locations in which an onset is occurring. Direction determination during these onset intervals enables discovery of the onsetting sound's azimuthal angle even in the presence of competing sounds. We compare the results from cross-correlation and novel spike based techniques for determining azimuthal angle during these onset intervals. We note that the system is suitable for real-time implementation.

Stochastic resonance (SR) is a phenomenon in which the response of a nonlinear system to a subthreshold information-bearing signal is optimized by the presence of noise. By considering a nonlinear system (network of leaky integrate-and-fire (LIF) neurons) that captures the functional dynamics of neuronal firing, we demonstrate that sensory neurons could, in principle harness SR to optimize the detection and transmission of weak stimuli. We have previously characterized this effect by use of signal-to-noise ratio (SNR). Here in addition to SNR, we apply an entropy-based measure (Fisher information) and compare the two measures of quantifying SR. We also discuss the performance of these two SR measures in a full precision floating point model simulated in Java and in a precision limited integer model simulated on a field programmable gate array (FPGA). We report in this study that stochastic resonance which is mainly associated with floating point implementations is possible in both a single LIF neuron and a network of LIF neurons implemented on lower resolution integer based digital hardware. We also report that such a network can improve the SNR and Fisher information of the output over a single LIF neuron.

We present a neurally inspired technique for detecting onsets in sound. The outputs from a cochlea-like filter are spike coded, in a way similar to the auditory nerve. These AN-like spikes are presented to leaky integrate-and-fire (LIF) neurons through a depressing synapse. The spike outputs from these are then processed by another layer of LIF neurons. Onsets are detected with essentially zero latency. We present results from the TIMIT database.

A biologically inspired technique for detecting onsets in sound is presented. Outputs from a cochlea-like filter are spike coded, in a way similar to the auditory nerve (AN). These AN-like spikes are presented to a leaky integrate-and-fire neuron through a depressing synapse. Onsets are detected with essentially zero latency relative to these AN spikes. Onset detection results for a tone burst, musical sounds and the DARPA/NIST TIMIT speech corpus are presented.

We describe a design technique for neuromorphic engineering that exploits both Mead's notion of physical equivalence between transistors and ion channels and the intrinsic properties of the materials of which transistors are composed. This is in contrast to the "algorithmic" design technique in common use. A "physical" design technique allows us to emphasise the morphology of cells. We hope to use this in an exploration of the effect of cell morphology upon function.

We present the SPIKEII chip, an integrate-and-fire neural network chip with programmable synapses implemented in analogue VLSI. It is the successor to the SPIKEI chip. We describe the circuitry, and show some results using networks of integrate-and-fire neurons.

A simple laterally inhibited recurrent network that implementse xclusive-or is demonstrated. The network consists of two mutually inhibitory units with logistic output function, each receiving one external input and each connected to a simple threshold output unit. The mutually inhibitory units settle into a point attractor. We investigate the range of steepness of the logistic and the range of inhibitory weights for which the network can perform exclusive-or.

This paper investigates the constraints placed on some synchronized oscillator models by their underlying dynamics. Phase response graphs are used to determine the phase locking behaviors of three oscillator models. These results are compared with idealized phase response graphs for single phase and multiple phase systems. We find that all three oscillators studied are best suited to operate in a single phase system and that the requirements placed on oscillatory models for operation in a multiple phase system are not compatible with the underlying dynamics of oscillatory behavior for these types of oscillator models.

A notation for the specification of neural nets is proposed. The aim is to produce a simple mathematical framework for use in specifying neural nets essentially by defining their transfer functions and connections. Nets are specified as interacting processing elements (nodes), communicating via instant links. Dynamics and adaptation are defined at the processing elements themselves, and all interaction is explicitly specified by directed arcs. Specifications can be built up hierarchically by turning a specification into a generator for a node, or they can be developed top-down. The use of the system is illustrated.>

. 1990) can support an error-correcting learning rule. The rule increases weights when both pre- and postsynaptic units are highly active, and decreases them when pre-synaptic activity is high and postsynaptic activation is less than the threshold for weight increment but greater than a lower threshold. We show that this rule corrects false positive outputs in feedforward associative memory, that in an appropriate opponent-unit architecture it corrects misses, and that it performs better than the optimal Hebbian learning rule reported by Willshaw and Dayan (1990).