An Adaptive Kernels Layer for Deep Neural Networks Based on Spectral Analysis for Image Applications
Author:
Al Shoura Tariq1ORCID, Leung Henry1, Balaji Bhashyam2ORCID
Affiliation:
1. Department of Electrical and Software Engineering, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4, Canada 2. Radar Sensing and Exploitation Section, Defence Research and Development Canada, Ottawa, ON K1A 0Z4, Canada
Abstract
As the pixel resolution of imaging equipment has grown larger, the images’ sizes and the number of pixels used to represent objects in images have increased accordingly, exposing an issue when dealing with larger images using the traditional deep learning models and methods, as they typically employ mechanisms such as increasing the models’ depth, which, while suitable for applications that have to be spatially invariant, such as image classification, causes issues for applications that relies on the location of the different features within the images such as object localization and change detection. This paper proposes an adaptive convolutional kernels layer (AKL) as an architecture that adjusts dynamically to images’ sizes in order to extract comparable spectral information from images of different sizes, improving the features’ spatial resolution without sacrificing the local receptive field (LRF) for various image applications, specifically those that are sensitive to objects and features locations, using the definition of Fourier transform and the relation between spectral analysis and convolution kernels. The proposed method is then tested using a Monte Carlo simulation to evaluate its performance in spectral information coverage across images of various sizes, validating its ability to maintain coverage of a ratio of the spectral domain with a variation of around 20% of the desired coverage ratio. Finally, the AKL is validated for various image applications compared to other architectures such as Inception and VGG, demonstrating its capability to match Inception v4 in image classification applications, and outperforms it as images grow larger, up to a 30% increase in accuracy in object localization for the same number of parameters.
Funder
Natural Science and Engineering Research Council Defence Research and Development Canada
Subject
Electrical and Electronic Engineering,Biochemistry,Instrumentation,Atomic and Molecular Physics, and Optics,Analytical Chemistry
Reference32 articles.
1. Building change detection in multitemporal very high resolution SAR images;Marin;IEEE Trans. Geosci. Remote Sens.,2014 2. Building Change Detection in VHR SAR Images via Unsupervised Deep Transcoding;Saha;IEEE Trans. Geosci. Remote Sens.,2021 3. Imagenet classification with deep convolutional neural networks;Krizhevsky;Adv. Neural Inf. Process. Syst.,2012 4. Simonyan, K., and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv. 5. Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2015, January 7–12). Going deeper with convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Boston, MA, USA.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. SEPE Dataset: 8K Video Sequences and Images for Analysis and Development;Proceedings of the 14th Conference on ACM Multimedia Systems;2023-06-07
|
|