| --- |
| license: agpl-3.0 |
| library_name: onnx |
| pipeline_tag: object-detection |
| tags: |
| - yolov8 |
| - onnx |
| - object-detection |
| - coco |
| - aidge |
| --- |
| |
| # YOLOv8m ONNX |
|
|
| COCO-pretrained YOLOv8m model exported from Ultralytics to ONNX for Aidge validation. |
|
|
| ## Model |
|
|
| - Source model: Ultralytics YOLOv8m |
| - Task: object detection |
| - Dataset: COCO |
| - Export format: ONNX |
| - ONNX opset: 17 |
| - Input: `images`, shape `[1, 3, 640, 640]` |
| - Output: `output0`, shape `[1, 84, 8400]` |
|
|
| ## Usage in Aidge |
|
|
| This model is used to validate YOLOv8 CPU import and inference in Aidge through the ONNX import path. |
|
|
| The static `640x640` export is used for model-zoo testing because it has fixed input and output shapes. |
|
|
| ## License |
|
|
| The original YOLOv8 model is provided by Ultralytics under the AGPL-3.0 license. |
|
|
