Abstract
We present a new parallel algorithm for image feature extraction. which uses a distance function based on the LZ-complexity of the string representation of the two images. An input image is represented by a feature vector whose components are the distance values between its parts (sub-images) and a set of prototypes. The algorithm is highly scalable and computes these values in parallel. It is implemented on a massively parallel graphics processing unit (GPU) with several thousands of cores which yields a three order of magnitude reduction in time for processing the images. Given a corpus of input images the algorithm produces labeled cases that can be used by any supervised or unsupervised learning algorithm to learn image classification or image clustering. A main advantage is the lack of need for any image processing or image analysis; the user only once defines image-features through a simple basic process of choosing a few small images that serve as prototypes. Results for several image classification problems are presented.
Keywords
- Graphic Processing Unit
- Input Image
- Image Classification
- Image Feature Extraction
- Unsupervised Learning Algorithm
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Fei-Fei, L., Fergus, R., Perona, P.: Learning generative visual models from few training examples: an incremental bayesian approach tested on 101 object categories (2004)
Lazebnik, S., Schmid, C., Ponce, J.: A sparse texture representation using local affine regions. IEEE Trans. Pattern Anal. Mach. Intell. 27(8), 1265–1278 (2005)
Chester, U., Ratsaby, J.: Universal distance measure for images. In: Proceedings of the 27th IEEE Convention of Electrical Electronics Engineers in Israel (IEEEI 2012), pp. 1–4. Eilat, Israel, 14–17 November 2012
Sayood, K., Otu, H.H.: A new sequence distance measure for phylogenetic tree construction. Bioinformatics 19(16), 2122–2130 (2003)
Ziv, J., Lempel, A.: On the complexity of finite sequences. IEEE Trans. Inf. Theory 22(3), 75–81 (1976)
Chester, U., Ratsaby, J.: Machine learning for image classification and clustering using a universal distance measure. In: Brisaboa, N., Pedreira, O., Zezula, P. (eds.) SISAP 2013. LNCS, vol. 8199, pp. 59–72. Springer, Heidelberg (2013)
Belousov, A., Ratsaby, J.: Massively parallel computations of the LZ-complexity of strings. In: Proceedings of the 28th IEEE Convention of Electrical and Electronics Engineers in Israel (IEEEI 2014), pp. 1–5. Eilat, 3–5 December 2014
Hall, M., Frank, E., Holmes, G., Pfahringer, B., Reutemann, P., Witten, I.H.: The WEKA data mining software: an update. SIGKDD Explor. 11(1), 10–18 (2009)
Acknowledgement
We acknowledge the support of the nVIDIA corporation for their donation of GPU hardware.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Belousov, A., Ratsaby, J. (2015). A Parallel Distributed Processing Algorithm for Image Feature Extraction. In: Fromont, E., De Bie, T., van Leeuwen, M. (eds) Advances in Intelligent Data Analysis XIV. IDA 2015. Lecture Notes in Computer Science(), vol 9385. Springer, Cham. https://doi.org/10.1007/978-3-319-24465-5_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-24465-5_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-24464-8
Online ISBN: 978-3-319-24465-5
eBook Packages: Computer ScienceComputer Science (R0)