direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments


Search for publications

Journal Publications by GRK members

A Neuromorphic Model of Dual Pathway Odour Classification
Citation key Haeusler20100
Author Häusler, C., Nawrot, M. P., and Schmuker, M.
Year 2010
DOI 10.3389/conf.fncom.2010.51.00043
Journal Front. Comput. Neurosci. Conference Abstract: Bernstein Conference on Computational Neuroscience
Volume 4
Month September
Abstract Insects can identify odorant stimuli in a fast and highly reliable manner. This is evident from a large body of behavioral discrimination experiments, e.g. in the honeybee (Giurfa 2007). The neuronal circuits involved in odor discrimination are well described on the structural level. We investigate the functional implications at the circuit level in order to improve our understanding of insect olfaction at each processing step. Our ultimate goal is to employ neural working principles for processing of chemical sensor data in technical applications.Here, we show that strong lateral inhibitory connections found in the insect antennal lobe (AL) facilitate more accurate odour classification by increasing the linear separability of presented stimuli. Our findings are based on a spiking neural network extension of prior work with rate coding in the AL (Schmuker et al. 2007). By leveraging the pyNN neural network modeling language (Eppler et al 2009), we are able to run simulations using well established neuronal simulators such as Neuron and Nest along with neuromorphic hardware that employs up to several hundred analog implementations of Integrate and Fire neurons and their synaptic interconnections (Brüderle et al. 2009).As an extension of our current work, we will also investigate a spiking network model of the dual olfactory pathways in the AL of Hymenoptera (bees and ants). The first is characterised by a narrow tuning to different odours whilst being largely concentration invariant. The second is less selective toward odour identity but responds strongly to different odour concentrations (Yamagata et al. 2009). These complementary features imply a functional diversification, which possibly results in an improved odor discriminability in bees and ants over other insect species.
Link to original publication Download Bibtex entry

To top

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions