Downloads · 30 days
0
basketball-deim-m-640
ortizeg/basketball-deim-m-640
basketball-deim-m-640 is a object detection model from ortizeg. Use it when you need objects located in an image. It is set up for onnx. The card lists the license as apache-2.0.
DEIM (D-FINE variant) medium transformer detector fine-tuned on the basketball dataset at 640x640. Smallest of the 7-model final-comparison roster (19.3M params) and the top Apache-2.0 scorer on the harness 10-class t…
Access
Public
Updated Jul 30, 2026
Repo size
77.7 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.onnx77.7 MB · 100%
At a glance
- Task
- Object Detection
- Library
- onnx
- License
- apache-2.0
- Access
- Public
- Created
- Jul 30, 2026
- Updated
- Jul 30, 2026
- SHA
- 1b4f378b
Model card
From the Hugging Face model README
deim-m-640
DEIM (D-FINE variant) medium transformer detector fine-tuned on the basketball dataset at 640x640. Smallest of the 7-model final-comparison roster (19.3M params) and the top Apache-2.0 scorer on the harness 10-class test slice.
Metrics
Measured on basketball-player-detection-3 (test split), via the object-detection-eval harness.
| Metric | 5-class | 10-class |
|---|---|---|
| mAP@50:95 | 0.686 | 0.643 |
| mAP@50 | 0.942 | 0.858 |
Preprocessing
- Resize:
square(alignment:none) - Normalize:
div255 - Channel order:
RGB - Input shape:
(3, 640, 640)(float32)
Provenance
- Source repo: https://github.com/Intellindust-AI-Lab/DEIM
- Training config:
docs/provenance/configs/deim_m/deim_basketball_m.yml - Hardware: vast.ai A100 80GB (single GPU)
- Command:
see docs/provenance/training-runs.md (DEIM-D-FINE-M @640 row); ckpt best_stg1.pth
Usage
This ONNX file is one of the 7-model roster benchmarked in object-detection-eval. Load it through the registry for verified, hash-checked download and the exact preprocessing recipe above:
from object_detection_eval.registry import ModelRegistry, download_weights
registry = ModelRegistry.from_directory("registry")
card = registry.get("deim-m-640")
weights_path = download_weights(card)
- Task
- Object Detection
- Library
- onnx
- License
- apache-2.0
- Created
- Jul 30, 2026
- Updated
- Jul 30, 2026