I'm a beginner in computer vision. I want to know which option among the following two can get better accuracy of image classification.

  1. SIFT features + SVM
  2. Bag-of-visual-words features + SVM

Here's a reference: https://www.mathworks.com/help/vision/ug/image-classification-with-bag-of-visual-words.html.

  • $\begingroup$ Basically, deep learning is the current method for object classification. Pretty much all classic areas are dead. $\endgroup$ – FourierFlux Sep 26 at 0:22

Bag-of-visual words (BOVW) was classicly used in computer vision before the introduction of neural networks or some more advanced classical techniques, such us VLAD or Fisher Vectors. In any case, it is a good technique to use, but it is not the state-of-the-art today, and I won't recommend you use it for a real-life project.

| improve this answer | |
  • $\begingroup$ Thanks for your advice! As I understand it, the BoW first apply knn to create a histogram vector and then apply svm to classify. I'm not sure such a technique is called "vector quantization" or "two stage classification". And I'm curious about that if such a technique improves the accuracy of classification. $\endgroup$ – Jimmy116 Feb 10 at 8:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.