Vombit/yolo12s_cs2

Counter Strike 2 players detector

Supported Labels

[ 'c', 'ch', 't', 'th' ]

All models in this series

• yolo12n_cs2 (~5mb)
• yolo12s_cs2 (~18mb)
• yolo12m_cs2 (~39mb)
• yolo12l_cs2 (~51mb)
• yolo12x_cs2 (~113mb)

How to use

# load Yolo
from ultralytics import YOLO

# Load a pretrained YOLO model
model = YOLO(r'weights\yolo**_cs2.pt')

# Run inference on 'image.png' with arguments
model.predict(
    'image.png',
    save=True,
    device=0
    )

Predict info

Ultralytics 8.3.105 🚀 Python-3.12.0 torch-2.8.0.dev20250408+cu128 CUDA:0 (NVIDIA GeForce RTX 4060, 8187MiB)

• yolo12s_cs2_fp16.engine (384x640 5 ts, 5 ths, 6.2ms)
• yolo12s_cs2.engine (384x640 5 ts, 5 ths, 7.1ms)
• yolo12s_cs2_fp16.onnx (640x640 5 ts, 5 ths, 101.4ms)
• yolo12s_cs2.onnx (384x640 5 ts, 5 ths, 52.8ms)
• yolo12s_cs2.pt (384x640 5 ts, 5 ths, 62.3ms)

Dataset info

Data from over 150 games, where the footage has been tagged in detail.

Train info

The training took place 100 epochs.

You can also support me with a cup of coffee: donate