Repository navigation
How does the Yolov8 handle multiple bounding boxes in the dataset? #56
Description
I have changed my dataset into xml files for each image, however I have multiple tags as I have multiple bounding boxes in my dataset. When running the tutorial 11, the train_yolo.py has some problems. The dataset loads correctly, [None, path to annotations].
Has anyone found a way to train the model for multiple bounding boxes?
Activity
pythonlessons commentedon Jun 6, 2024
Give me your dataset and code you try to run. We'll solve it
Rajaro72 commentedon Jun 6, 2024
Thank you! I tried two attempts, one was to change the VOCannotationsclass to read the CSV files and another attempt which converted the dataset into xml files for each image. I will attach the code files to this comment and here is the link to the dataset:https://www.kaggle.com/datasets/robikscube/textocr-text-extraction-from-images-dataset?select=train_val_images
XML_attempt.zip
CSV_attempt.zip