Skip to Main Content
This paper presents an effective combination of SIFT features and random features. For the combined feature patches extracted from images we then adopt the PCA transformation to reduce the dimensionality of their feature vectors. And the reduced vectors are categorized by Gaussian mixture models (GMMs) in witch the mixture weights are adjusted iteratively using gradient descent. We experiment on Caltech datasets using this enhanced method, and the results comparing with several other methods show that the combination of salient feature vectors and GMM gives a much better improvement in object recognition.