CoreML export. How to parse MLMultiArray

[obohrer]

Search before asking

• I have searched the YOLOv5 issues and discussions and found no similar questions.

Question

Hi, I'm relatively new to this domain and currently experimenting with running CoreML to detect objects on IOS.
I've managed to train a custom model using yolov5 and tested the weights using a webcam as source. It does work 🚀.
However, when exporting the model to coreml using export.py it has been a struggle.
The exported model has the input dimension 480x640, so far so good.
However, the outputs are the following var_875, var_860.
During detection CoreML returns 2 matrices :[1 x 3 x 80 x 60 x 6] and [1 x 3 x 40 x 30 x 6]
It seems that 80 x 60 and 40 x 30 are the same ratio as the dimension (*8,*16 multiplier)
Is there documentation on how to parse such outputs?

So far I've found many functions to parse those MLMultiArrayValue but none of them are producing correct scores or bounding boxes.

Any pointers on what those dimensions means and how to parse it would be greatly appreciated and happy to contribute back with code snippets / improve the exporter (eg name the outputs correctly)

Additional

No response

[github-actions]

👋 Hello @obohrer, 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 screenshots and minimum viable code to reproduce your issue, otherwise we can not help you.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset images, training logs, screenshots, and a public link to online W&B logging if available.

For business inquiries or professional support requests please visit https://ultralytics.com or email Glenn Jocher at [email protected].

Requirements

Python>=3.6.0 with all requirements.txt installed including PyTorch>=1.7. To get started:

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

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):

• Google Colab and Kaggle 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 (train.py), validation (val.py), inference (detect.py) and export (export.py) on MacOS, Windows, and Ubuntu every 24 hours and on every commit.

[obohrer]

This repository has been useful: https://github.com/dbsystel/yolov5-coreml-tools
Had to tweak the coremltools dependency and skip one of the outputNames in yolov5-coreml-tools.
So far it seems that the bounding boxes are correct, same as predictions.
Only issue I'm seeing on iOS now is the confidence always show 100% (same as with models created with Create ML when there is one class)

[glenn-jocher]

@obohrer good news 😃! Your original issue may now be fixed ✅ in PR #6195. This PR adds support for YOLOv5 CoreML inference.

!python export.py --weights yolov5s.pt --include coreml  # CoreML export
!python detect.py --weights yolov5s.mlmodel  # CoreML inference (MacOS-only)
!python val.py --weights yolov5s.mlmodle  # CoreML validation (MacOS-only)

model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s.mlmodel')  # CoreML PyTorch Hub model

To receive this update:

• Git – git pull from within your yolov5/ directory or git clone https://github.com/ultralytics/yolov5 again
• PyTorch Hub – Force-reload model = torch.hub.load('ultralytics/yolov5', 'yolov5s', force_reload=True)
• Notebooks – View updated notebooks
• Docker – sudo docker pull ultralytics/yolov5:latest to update your image

Thank you for spotting this issue and informing us of the problem. Please let us know if this update resolves the issue for you, and feel free to inform us of any other issues you discover or feature requests that come to mind. Happy trainings with YOLOv5 🚀!

[tcollins590]

This repository has been useful: https://github.com/dbsystel/yolov5-coreml-tools Had to tweak the coremltools dependency and skip one of the outputNames in yolov5-coreml-tools. So far it seems that the bounding boxes are correct, same as predictions. Only issue I'm seeing on iOS now is the confidence always show 100% (same as with models created with Create ML when there is one class)

Would you be able to share your updated dependency and which outputNames you skipped? Thank you