How to use hierarchical classifier to determine unknown class

[raihan (Azizul Hakim)]

I have trained a CNN model for multiclass classification. But the unknown images are classified as one of the classes. I need to classify the random images as “other” class. Searching online I have found that a hierarchical classifier is a good way to solve this problem. Initially, an SVM or CNN classifier can be used to detect the outliers. Then I can use my CNN model for multiclass classification.
However, I have no idea how to combine 2 classification models to make this work. Can anyone help me with the technical details?

[mrfabulous1]

Hi raihan hope all is well!
Unfortunately I can’t solve your issue, I am currently just completing Lesson 4, however I am interested in your problem. It would be great if you could use the thread below as there are more people trying to solve this problem. Someone may have an Idea that can help you.

Handle data that belongs to classes not seen in training or testing

Let’s say I have trained an image recogniser to classify the content of an image as either a dog, a cat, a rat or an elephant. I put my model into production and it is for some reason fed with images of humans. The probabilities from the model suggests that there is 70% probability that an image of a human is an image of an elephant. I would rather like the model to say the input is not like any of the classes it was trained on. What is the go-to-way of handling this problem? I don’t have resou…

cheers mrfabulous1

[raihan (Azizul Hakim)]

Thank you @mrfabulous1 for the help. I got some useful information.