ABSTRACT
Maritime data uniquely challenges imagery analysis. Such data suffers from degradation, limited samples, and varied formats. To this end, the Hypercube-based NeuroEvolution of Augmenting Topologies (HyperNEAT) approach is investigated in addressing such challenges for classifying maritime vessels in a satellite imagery data set. The results show that HyperNEAT learns to extract features that allows better classification than those from Principal Component Analysis (PCA) and robust to differences in presentation of data. Furthermore, HyperNEAT enables a unique capability to scale trained solutions to different image resolutions.
- H. Buck, E. Sharghi, C. Guilas, J. Stastny, W. Morgart, B. Schalcosky, and K. Pifko. Enhanced ship detection from overhead imagery. In SPIE Def. and Sec. Symp., volume 6945. Inte. Soc. for Optics and Phot., April 2008.Google ScholarCross Ref
- J. Gauci and K. O. Stanley. Autonomous evolution of topographic regularities in artificial neural networks. Neural Computation, page 38, 2010. Google ScholarDigital Library
- J. Harguess*, S. Parameswaran*, K. Rainey*, and J. Stastny. Vessel classification in overhead satellite imagery using learned dictionaries. In SPIE Optical Eng.Google Scholar
- Apps.. Inter. Soc. for Optics and Phot., 2012.Google Scholar
- K. Rainey, S. Parameswaran, and J. Harguess. Maritime vessel recognition in degraded satellite imagery, 2014.Google Scholar
- P. A. Szerlip, G. Morse, J. K. Pugh, and K. O. Stanley. Unsupervised feature learning through divergent discriminative feature accumulation. arXiv:1406.1833v2, abs/1406.1833, 2014.Google Scholar
- P. Verbancsics and J. Harguess. Deep learning through generative and developmental system. In Proc. of the Gen. and Ev. Comp. Conf., pg. 103. ACM, 2014. Google ScholarDigital Library
- P. Verbancsics and J. Harguess. Image classification using generative neuroevolution for deep learning. In IEEE Winter Conf. on Apps. of Comp. Vision, Waikola, HI, January 2015. Google ScholarDigital Library
- P. Verbancsics and K. O. Stanley. Evolving static representations for task transfer. Journal of Machine Learning Research, 11, 2010. Google ScholarDigital Library
- P. Verbancsics and K. O. Stanley. Constraining connectivity to encourage modularity in hyperneat. In Proc. of the Gen. and Ev. Comp. Conf., New York, NY, 2011. ACM Press. Google ScholarDigital Library
Index Terms
- Classifying Maritime Vessels from Satellite Imagery with HyperNEAT
Recommendations
Constraining connectivity to encourage modularity in HyperNEAT
GECCO '11: Proceedings of the 13th annual conference on Genetic and evolutionary computationA challenging goal of generative and developmental systems (GDS) is to effectively evolve neural networks as complex and capable as those found in nature. Two key properties of neural structures in nature are regularity and modularity. While HyperNEAT ...
Investigating whether hyperNEAT produces modular neural networks
GECCO '10: Proceedings of the 12th annual conference on Genetic and evolutionary computationHyperNEAT represents a class of neuroevolutionary algorithms that captures some of the power of natural development with a computationally efficient high-level abstraction of development. This class of algorithms is intended to provide many of the ...
An enhanced hypercube-based encoding for evolving the placement, density, and connectivity of neurons
Intelligence in nature is the product of living brains, which are themselves the product of natural evolution. Although researchers in the field of neuroevolution NE attempt to recapitulate this process, artificial neural networks ANNs so far evolved ...
Comments