Which classifier is best for image classification?

Which classifier is best for image classification

Answer 1

[Abhishek Ballaney]

https://in.mathworks.com/help/nnet/deep-learning-image-classification.html

Answer 2

[Krishna]

Hey,

When selecting the best networks for image classification, the choice depends on your specific application needs. For instance, if achieving the highest accuracy is your goal and you’re not concerned about the number of parameters, you might consider using DenseNets or Vision Transformers. On the other hand, if you need a lightweight classifier for resource-constrained environments, MobileNets or EfficientNets would be more appropriate. The ideal network will vary based on the application you’re developing.

Please review the following documentation to learn more about various image classifiers,

https://www.geeksforgeeks.org/top-pre-trained-models-for-image-classification/

Also please go through the following documentation to learn more on how to do transfer learning in matlab to build your own image classifer,

https://in.mathworks.com/help/deeplearning/gs/classify-image-using-pretrained-network.html

This example shows in detail on how to implement pre-trained googlenet in matlab. You can follow this workflow for any other pre-trained network supported by MATLAB.

Hope this helps.