I have a data set that consists of images. I am trying to perform multi-label classification on this data set. But the training labels consist of too many labels which are CSV file format. Now I find it a little difficult on how to utilizes these labels in CSV file during the course of training. I read some blogs; the author suggested performing one-hot encoding on these labels. Now I am stuck on how to proceed. Below is the screenshot of the label.

screen shot

  • 1
    $\begingroup$ May I ask what task or you trying to do ? Is it multi class detection on an image ? like saying 'there is a cat, a dog and a bike on this image'. If it is, then the way to proceed is by creating a label vector [0 ,0 ,0,1, 1, 0, 0, 1] with ones on all the classes present on the image and zero if the class is not there. than use a regular classification architecture, but instead of using softmax function, use sigmoid function on the last layer. I can't remember which loss function is used tho for multi-class classification $\endgroup$
    – Ubikuity
    Jun 27, 2021 at 13:12
  • $\begingroup$ I am performing multi-label classification on the image $\endgroup$ Jun 27, 2021 at 13:31
  • $\begingroup$ Sry about the late answer, as I said, I think you should just create label vector that are [0, 0, 0, 1, 1, 0], with 1 being the classes in the image, and 0 the classes not in the image $\endgroup$
    – Ubikuity
    Jun 27, 2021 at 22:32


Your Answer

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

Browse other questions tagged or ask your own question.