A Hybrid Deep Learning Approach for Texture Analysis

Texture classification is a problem that has variousapplications such as remote sensing and forest speciesrecogni- tion. Solutions tend to be custom fit to the datasetused but fails to generalize. The Convolutional NeuralNetwork (CNN) in combination with Support Vector Machine(SVM) form a robust selection between powerful invariantfeature extractor and accurate classifier. The fusion ofexperts provides stability in classification rates amongdifferent datasets.

Paper

Similar papers

© 2026 NYSGPT2525 LLC