A question about bbox normalization about yolov5

[github-actions]

👋 Hello @doLei-2001, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Requirements

Python>=3.8.0 with all requirements.txt installed including PyTorch>=1.8. To get started:

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training, validation, inference, export and benchmarks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

Introducing YOLOv8 🚀

We're excited to announce the launch of our latest state-of-the-art (SOTA) object detection model for 2023 - YOLOv8 🚀!

Designed to be fast, accurate, and easy to use, YOLOv8 is an ideal choice for a wide range of object detection, image segmentation and image classification tasks. With YOLOv8, you'll be able to quickly and accurately detect objects in real-time, streamline your workflows, and achieve new levels of accuracy in your projects.

Check out our YOLOv8 Docs for details and get started with:

pip install ultralytics

[glenn-jocher]

Hello! Thanks for providing detailed information about the issue you're encountering. It seems like the error arises because the bounding box coordinates are not correctly normalized to the range (0, 1] after resizing and scaling transformations.

From your code snippet, it appears that you are normalizing the bounding box coordinates by dividing by target_size after scaling them. However, the error suggests that some values might still be out of the expected range. This could potentially be due to floating-point precision issues or initial values slightly outside the expected range before scaling.

To address this, you might want to ensure that the bounding box coordinates are strictly within the (0, 1] range before they are passed to the augmentation pipeline. Here's a modified version of your label updating loop that includes an additional check and correction for bounding box coordinates:

updated_labels = []
for label in labels4:
    class_id, x1, y1, x2, y2 = label

    # Scale bounding box coordinates
    new_x1 = x1 * scale_x
    new_y1 = y1 * scale_y
    new_x2 = x2 * scale_x
    new_y2 = y2 * scale_y

    # Normalize to [0, 1]
    new_x1 /= target_size
    new_y1 /= target_size
    new_x2 /= target_size
    new_y2 /= target_size

    # Ensure coordinates are within (0, 1]
    new_x1 = max(1e-5, min(new_x1, 1))
    new_y1 = max(1e-5, min(new_y1, 1))
    new_x2 = max(1e-5, min(new_x2, 1))
    new_y2 = max(1e-5, min(new_y2, 1))

    updated_labels.append([class_id, new_x1, new_y1, new_x2, new_y2])

Please make sure that the initial bounding box coordinates (x1, y1, x2, y2) are within the valid range before transformation. If the issue persists, you might want to add debug prints to check the values of coordinates at each step.

Let me know if this helps or if you have any more questions!