Autonomous_driving_application_Car_detection, indices[1] = 1 is not in [0,1) failure

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Hello, I’m stuck on Autonomous_driving_application_Car_detection with the error “InvalidArgumentError: indices[0] = 1 is not in [0, 1) [Op:GatherV2]”.

The error is happening at line of code “nms_indices.append(tf.gather(nms_indices_label,selected_indices))”. I found a different post with a very similar question here: Getting error in yolo_non_max_suppression function - #2 by balaji.ambresh

But still can’t figure this out. Swapping the order of parameters in tf.gather changes when the error occurs. It changes whether the error happens at the first or the 4th iteration of the for loop.

Any thoughts on how to proceed?

Thank you!

1 Like
3

This is a very common error and you can find lots of threads about it by searching for yolo_non_max_suppression. But did you read all the way through the thread you linked above? E.g. here’s a later post on that thread that covers more issues. And even later on that thread, here’s a very comprehensive post from mentor Raymond. Definitely worth a look, but you will need to have your thinking cap firmly affixed and your morning coffee already on board! :nerd_face:

2 Likes
4

Probably not, since sharing a programming assignment notebook is not allowed by the Code of Conduct.

5

Yup, my problem was covered in that thread after all. I was passing boxes and scores to nms function instead of boxes_label and scores_label.

Thank you!

2 Likes