huangrm/yolo-detector

ATEC2026 YOLO Object Detector

YOLOv8-nano model fine-tuned for detecting objects in the ATEC2026 Simulation Challenge.

Model Details

• Base model: YOLOv8-nano (yolo26n.pt)
• Training epochs: 100
• Input size: 640x640
• Classes: banana, box (sugar), mustard (3 classes)

Usage

from ultralytics import YOLO

model = YOLO("best.pt")
results = model.predict("image.png", conf=0.25)

Training

python scripts/train.py --data datasets/auto_collect/yolo_dataset/dataset.yaml --epochs 100

Dataset

Trained on atec2026-object-detection — 1000 simulated RGB images with bounding box annotations.