YOLOv8n

YOLOv8n is the smallest and fastest lightweight vision model in the YOLOv8 series released by Ultralytics. Based on an advanced deep learning architecture, it delivers excellent real-time detection performance with very low compute cost, making it a preferred choice for edge and mobile deployments.

• Key features: Supports high-accuracy real-time object detection, instance segmentation, image classification, and pose estimation (keypoint detection).
• Version notes: This example uses YOLOv8n (Nano). As a lightweight benchmark in the family, it achieves very high FPS with minimal parameters, greatly reducing hardware requirements while maintaining mainstream detection accuracy. It is a strong choice that balances real-time responsiveness and ease of deployment.

Environment setup

You need to set up the environment in advance.

• Environment setup
• AI Model Hub

Quick start

Download model files

O6 / O6N

cd ai_model_hub_25_Q3/models/ComputeVision/Object_Detection/onnx_yolov8_n
wget https://www.modelscope.cn/models/cix/ai_model_hub_25_Q3/resolve/master/models/ComputeVision/Object_Detection/onnx_yolov8_n/yolov8n.cix

Test the model

info

Activate the virtual environment before running.

O6 / O6N

python3 inference_npu.py

Full conversion workflow

Download model files

Linux PC

cd ai_model_hub_25_Q3/models/ComputeVision/Object_Detection/onnx_yolov8_n/model
wget https://www.modelscope.cn/models/cix/ai_model_hub_25_Q3/resolve/master/models/ComputeVision/Object_Detection/onnx_yolov8_n/model/yolov8n.onnx

Project structure

├── cfg
├── datasets
├── inference_npu.py
├── inference_onnx.py
├── model
├── ReadMe.md
├── test_data
└── yolov8n.cix

Quantize and convert the model

Linux PC

cd ..
cixbuild cfg/yolov8_nbuild.cfg

Copy to device

After conversion, copy the .cix model files to the device.

Test inference on the host

Run the inference script

Linux PC

python3 inference_onnx.py

Host inference output

Linux PC

$ python3 inference_onnx.py
100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 19.24it/s]

Deploy on NPU

Run the inference script

O6 / O6N

python3 inference_npu.py

Inference output

O6 / O6N

$ python3 inference_npu.py
npu: noe_init_context success
npu: noe_load_graph success
Input tensor count is 1.
Output tensor count is 1.
npu: noe_create_job success
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 18.07it/s]
npu: noe_clean_job success
npu: noe_unload_graph success
npu: noe_deinit_context success