Hybrid Semi-Parallel Deep Neural Networks (SPDNN) – example methodologies and use cases
MetadataShow full item record
This item's downloads: 10 (view details)
Corcoran, Peter, & Bazrafkkan, Shabab. (2018). Hybrid Semi-Parallel Deep Neural Networks (SPDNN) – example methodologies and use cases. Paper presented at the Embedded Vision Summit, Santa Clara Convention Center, Silicon Valley, Santa Clara, California, 22-23 May.
Deep neural networks (DNNs) are typically trained on specific datasets, optimized with particular discriminating capabilities. Often several different DNN topologies are developed solving closely related aspects of a computer vision problem. But to utilize these topologies together, leveraging their individual discriminating capabilities, requires implementing each DNN separately, increasing the cost of practical solutions. In this talk, a methodology to merge multiple deep networks using graph contraction is developed. This provides a single network topology, achieving significant reduction in size over the individual networks. More significantly, this merged SPDNN network can be re-trained across the combined datasets used to train the original networks, improving its accuracy over the original networks. The result is a single network that is more generic, but with equivalent – or often enhanced – performance over a wider range of input data. Examples of several problems in contemporary computer vision are solved using SPDNNs. These include significantly improving segmentation accuracy of eye-iris regions (a key component of iris biometric authentication) and mapping depth from monocular images, demonstrating equivalent performance to stereo depth mapping.
This item is available under the Attribution-NonCommercial-NoDerivs 3.0 Ireland. No item may be reproduced for commercial purposes. Please refer to the publisher's URL where this is made available, or to notes contained in the item itself. Other terms may apply.
The following license files are associated with this item: