Compressed Domain Image Classification Using a Dynamic-Rate Neural Network

Simple item page
dc.citation.firstpage217711
dc.citation.journalTitleIEEE Access
dc.citation.lastpage217722
dc.citation.volumeNumber8
dc.contributor.authorXu, Y.
dc.contributor.authorLiu, W.
dc.contributor.authorKelly, K.F.
dc.date.accessioned2021-02-10T14:04:12Z
dc.date.available2021-02-10T14:04:12Z
dc.date.issued2020
dc.description.abstractCompressed domain image classification performs classification directly on compressive measurements acquired from the single-pixel camera, bypassing the image reconstruction step. It is of great importance for extending high-speed object detection and classification beyond the visible spectrum in a cost-effective manner especially for resource-limited platforms. Previous neural network methods require training a dedicated neural network for each different measurement rate (MR), which is costly in computation and storage. In this work, we develop an efficient training scheme that provides a neural network with dynamic-rate property, where a single neural network is capable of classifying over any MR within the range of interest with a given sensing matrix. This training scheme uses only a few selected MRs for training and the trained neural network is valid over the full range of MRs of interest. We demonstrate the performance of the dynamic-rate neural network on datasets of MNIST, CIFAR-10, Fashion-MNIST, COIL-100, and show that it generates approximately equal performance at each MR as that of a single-rate neural network valid only for one MR. Robustness to noise of the dynamic-rate model is also demonstrated. The dynamic-rate training scheme can be regarded as a general approach compatible with different types of sensing matrices, various neural network architectures, and is a valuable step towards wider adoption of compressive inference techniques and other compressive sensing related tasks via neural networks.
dc.identifier.citationXu, Y., Liu, W. and Kelly, K.F.. "Compressed Domain Image Classification Using a Dynamic-Rate Neural Network." <i>IEEE Access,</i> 8, (2020) IEEE: 217711-217722. https://doi.org/10.1109/ACCESS.2020.3041807.
dc.identifier.digital9274326
dc.identifier.doihttps://doi.org/10.1109/ACCESS.2020.3041807
dc.identifier.urihttps://hdl.handle.net/1911/109836
dc.language.isoeng
dc.publisherIEEE
dc.rightsThis article is licensed under a Creative Commons license
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.titleCompressed Domain Image Classification Using a Dynamic-Rate Neural Network
dc.typeJournal article
dc.type.dcmiText
dc.type.publicationpublisher version
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
9274326.pdf
Size:
1.82 MB
Format:
Adobe Portable Document Format
Download
Collections
Faculty Publications
ECE Publications