Multiple streams breaks once the model is converted to ONNX about ultralytics

[glenn-jocher]

@LordSizzz, I'm glad to hear that the solution worked for you! If you have any more questions or run into other issues, feel free to reach out. Happy coding! 🚀

[github-actions]

👋 Hello @LordSizzz, thank you for your interest in Ultralytics YOLOv8 🚀! We recommend a visit to the Docs for new users where you can find many Python and CLI usage examples and where many of the most common questions may already be answered.

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.

Join the vibrant Ultralytics Discord 🎧 community for real-time conversations and collaborations. This platform offers a perfect space to inquire, showcase your work, and connect with fellow Ultralytics users.

Install

Pip install the ultralytics package including all requirements in a Python>=3.8 environment with PyTorch>=1.8.

pip install ultralytics

Environments

YOLOv8 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 Ultralytics CI tests are currently passing. CI tests verify correct operation of all YOLOv8 Modes and Tasks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

[glenn-jocher]

@LordSizzz hello,

Thank you for providing detailed information about the issue you're encountering with ONNX and multiple RTSP streams. The error message you're seeing suggests that there's a mismatch in the expected input dimensions for your model. The ONNX model is expecting a single input stream (batch size of 1), but it seems to be receiving multiple streams simultaneously (batch size of 6).

This can occur if the model was exported with a specific batch size that doesn't match the number of streams you're trying to process. Here are a couple of steps you can take to resolve this:

1. Re-export the Model with Dynamic Axes: When exporting your model to ONNX, ensure that you enable dynamic axes to allow for variable batch sizes. This can be done by setting the dynamic=True parameter in the model.export() function. This will allow the model to handle varying numbers of input streams.

2. Adjust the Input Feed: Ensure that the input to the model is correctly batched according to the model's expected input dimensions. If your model expects a batch size of 1, you might need to process each stream individually or modify the model to accept the actual number of streams you intend to use.

Here's an example of how you might adjust the export code:

model.export(format="onnx", dynamic=True)

Please try these adjustments and let us know if the issue persists. We're here to help!

[LordSizzz]

@glenn-jocher thank you for the quick response, what you just mentioned whas indeed the issue, due to my lack of knowledge and the fact that it worked just before the conversion caused me to think that, so yes, adding ,dynamic=True in the export code did fix my problem, thank you for your help.