Yolo.h5 file/ setting up yoloV2 on my macbook

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

When creating a post, please add:

  • Week 3 must be added in the tags option of the post.

  • Link to the classroom item you are referring to: Lab: Autonomous_driving_application_Car_detection

  • I completed the lab on the coursera setup and then I tried to set the YOLOv2 on my macbook. I am using tensorflow 2.16.2 on jupyter to run my code. I cloned yolo v2 code from github. I downloaded yolo weights and yolo cfg files in model_data folder. I tried to generate yolo.h5 file using python3 yad2k.py model_data/yolov3.cfg model_data/yolov3.weights model_data/yolo.h5 but getting a lot of errors. I even changed a lot of code on yad2k.py to do away with errors but still it is not running and getting value errors now. Is there any easy way to do it. Please suggest. Thanks

2

To use a DLAI notebook locally, you have to install all of the same versions of the packages and tools that Coursera provides.

Otherwise, you have to solve all of the incompatibility problems that newer versions cause.

3

Thanks for your response.
I am finally able to set it up on my machine. However these are the some issues I am facing:

  1. Parameters mismatch:
    Coursera Lab"s parameters on loading the model -
    Total params: 50,983,561
    Trainable params: 50,962,889
    Non-trainable params: 20,672
    Parameters on my noterbook:
    Total params: 39,809,673
    Trainable params: 39,789,001
    Non-trainable params: 20,672
  2. Output not giving expected result:
    This is the output I got when ran the code in my notebook: “Found 0 boxes for
    images/dog.jpg”. The image of dog is displayed without a bounding box and no
    confidence score.

Can you please help me with the yad2k.py file and yolo.h5 so I can try running it. I am not able to find it in the lab’s folder.

Thanks

4

Sorry, I have no experience in installing these packages locally.

If you use the Forum search tool for those file names, you will probably find some information.

5

Hi @Sunil_Kapil
Maybe this post helps.

2 Likes