Questions about display a category of one image about pytorch_d-rise HOT 4 CLOSED

Thank you for sharing. I am interesting about this work. While I use the YOLOv3 https://github.com/ultralytics/yolov3/ as the object detector. I can generate the salience map when the image only have one object with a category, like the followings.

While it can't work when the image contain multiple objects of a categories like the persons in this image

The code compute the score I revised it as 'conf = max([iou(target_box, box)*conf], default=0)'

Could you give me some advice?

I applied on Yolo v3 and don't appear the problem you said.

from pytorch_d-rise.

@RuoyuChen10 could you share your code with me? My email is [email protected].

from pytorch_d-rise.

@RuoyuChen10 could you share your code with me? My email is [email protected].

sure, I will send it to you soon, and I will also upload it to my GitHub in the future: https://github.com/RuoyuChen10/objectdetection-saliency-maps

from pytorch_d-rise.

Sorry I missed this issue. Looks like it's already been solved, so I'll close this.

from pytorch_d-rise.