European Conference on Computer Vision · 2002 · 253 citations · 0 references
Image ClassificationImage AnalysisMachine VisionData ScienceEngineeringPattern RecognitionCategorizationVisual CategorizationObject CategorizationComputer ScienceComputer Vision
The authors introduce a bag‑of‑keypoints approach for generic visual categorization that generalizes across object‑class variations and compare two classifier implementations, Naive Bayes and SVM. The method vector‑quantizes affine‑invariant keypoint descriptors from image patches and applies Naive Bayes or SVM classifiers to simultaneously classify seven semantic categories. The approach is simple, computationally efficient, intrinsically invariant, and achieves robust accuracy in the presence of background clutter without using geometric information.
We present a novel method for generic visual categorization: the problem of identifying the object content of natural images while generalizing across variations inherent to the object class. This bag of keypoints method is based on vector quantization of affine invariant descriptors of image patches. We propose and compare two alternative implementations using different classifiers: Naive Bayes and SVM. The main advantages of the method are that it is simple, computationally efficient and intrinsically invariant. We present results for simultaneously classifying seven semantic visual categories. These results clearly demonstrate that the method is robust to background clutter and produces good categorization accuracy even without exploiting geometric information.