strinova-player-detection
YOLOv8 real-time player detection trained on Strinova gameplay footage.
Models
| Variant | Size | mAP@0.5 | Precision | Recall |
|---|---|---|---|---|
| YOLOv8n (nano) | 6MB | 97.7% | 96.6% | 92.3% |
| YOLOv8s (small) | 22MB | 95.2% | 92.5% | 92.3% |
| YOLOv8s TensorRT | 25MB | — | — | — |
| YOLOv8s TensorRT FP32 | 61MB | — | — | — |
Trained on RTX 4070 Ti. Dataset: 643 annotated images from Roboflow Universe.
Usage
from ultralytics import YOLO
model = YOLO("yolov8n.pt")
results = model.predict("image.jpg")