0

I'm currently working on Yolov3 and have spent the last two days trying to implement the Grade-CAM approach without success. At the end I link both github repositories I used.

Since I failed to create a heatmap with the approach I used before, I am looking for other ways to create a heatmap for a class and a picture. But so far I could not find any implementation how to do this.

Which approaches could I also pursue? Or which ideas should I still try?

Yolov3: https://github.com/zzh8829/yolov3-tf2 Grad-Cam: https://github.com/sicara/tf-explain

john-mueller
  • 107
  • 2
  • 9
  • What's your actual requirement? – Zabir Al Nazi May 10 '20 at 22:27
  • @zabir-al-nazi My actual need is to visually display the output of Yolov3 using a heat map. Depending on the class, it should be possible to display the relevant parts of a picture by "heat". I also work with Tensorflow 2. – john-mueller May 11 '20 at 07:16

1 Answers1

0

This notebook is great https://colab.research.google.com/drive/1CJQSVaXG1ceBQGFrTHnt-pt1xjO4KFDe. It's by the creator of Keras and it uses TF2.

Kh4zit
  • 2,629
  • 3
  • 13
  • 25
  • My problem is that what I have had with all the implementations of Grad-Cam is that the keras model can't be created in this step and I can't use Grad-Cam. The step I mean is : keras.model(input_laye, last_conv_layer.output) – john-mueller May 16 '20 at 09:47
  • You need to provide more context. Your explanation is not clear. – Kh4zit May 16 '20 at 15:13