Skip to main content

YOLO11 Pose

This document describes how to run YOLO11 Pose on NPU.

info

Refer to Model Zoo Download for the example.

YOLO11 Pose Example Directory Structure:

$ tree ./
./
├── CMakeLists.txt
├── convert_model
│   ├── config_yml.py
│   ├── convert_model_env.sh
│   ├── python
│   │   ├── onnx_extract.py
│   │   └── yolo11s-pose_640.txt
│   └── yolo11s-pose_9.txt
├── figures
│   ├── diff_img.png
│   └── out_yolo11_pose_pcq.png
├── main.cpp
├── model
│   └── COCO_train2014_000000500390.jpg
├── model_config.h
├── README.md
├── yolo11_pose_9_post.cpp
└── yolo11_pose_9_pre.cpp

Model Conversion​

Configure Virtual Environment​

X86 Linux PC
python -m venv .venv && source .venv/bin/activate
pip install ultralytics

Export ONNX Model​

X86 Linux PC
cd convert_model/python/
yolo export model=yolo11s-pose.pt format=onnx simplify=True dynamic=False opset=11 nms=False batch=1 device=cpu

Prune Model​

X86 Linux PC
python onnx_extract.py
mv ./yolo11s-pose_9.onnx ../
cd ..
X86 Linux PC
./convert_model_env.sh

Model Import/Quantization/Conversion​

You need to enter the container development environment first. Refer to the Create Container section in Model Zoo Download.

info

Different platforms use corresponding Docker images:

  • A733: ubuntu-npu:v2.0.10.1
  • T527: ubuntu-npu:v1.8.11
X86 Linux PC
docker exec -it model-zoo /bin/bash

After entering the container, navigate to the corresponding directory and run the script.

X86 Linux PC
cd /workspace/examples/yolo11_pose/convert_model/
X86 Linux PC
./pegasus_import.sh yolo11s-pose_9
./pegasus_quantize.sh yolo11s-pose_9 uint8 12
X86 Linux PC
./pegasus_export_ovx_nbg.sh yolo11s-pose_9 uint8 a733

The exported model files are stored in the ../model directory.

Compile Example​

Now you can compile the example. First exit the container, then execute the following command to compile the example.

First, you need to configure third-party libraries and cross-compilation toolchain.

info

You can skip this step if you have already configured third-party libraries and cross-compilation toolchain in other examples.

X86 Linux PC
cd ../../../3rdparty/opencv/
unzip opencv-4.9.0-aarch64-linux-sunxi-glibc.zip
cd ../../0-toolchains/

You need to manually download via this link first, then place it in 0-toolchains/ before executing the following command:

X86 Linux PC
tar -xvf gcc-arm-10.2-2020.11-x86_64-aarch64-none-linux-gnu.tar.xz
X86 Linux PC
cd ../examples/yolo11_pose/
X86 Linux PC
../build_linux.sh -t a733 -s debian11

Model Deployment​

After compilation, the example will be installed in the install directory. You can use scp to transfer it to the board.

Configure NPU Driver​

info

You can skip this step if you have already configured NPU driver in other examples.

Transfer the driver library to the board's lib directory via scp.

  • A733 corresponds to the common/lib_linux_aarch64/A733 directory
  • T527 corresponds to the common/lib_linux_aarch64/T527 directory

Then execute the following command to export to environment variables.

Radxa SBC
echo 'export LD_LIBRARY_PATH=$HOME/lib:$LD_LIBRARY_PATH' >> ~/.bashrc

Run Example​

After configuring the driver, you can run the example.

tip

For T527 platform, you need to first enable NPU by referring to the A5E's "Enable NPU on Board" documentation, then use the following command to grant the current user permission to use /dev/vipcore.

Radxa SBC
sudo chmod 777 /dev/vipcore
Radxa SBC
cd yolo11_pose_demo_linux_a733/
Radxa SBC
chmod +x ./yolo11_pose_demo_a733
./yolo11_pose_demo_a733 -nb model/yolo11s-pose_9_uint8_a733.nb -i model/COCO_train2014_000000500390.jpg

The running result is as follows:

$ ./yolo11_pose_demo_a733 -nb model/yolo11s-pose_9_uint8_a733.nb -i model/COCO_train2014_000000500390.jpg
model_file=model/yolo11s-pose_9_uint8_a733.nb, input=model/COCO_train2014_000000500390.jpg, loop_count=1, malloc_mbyte=10
VIPLite driver software version 2.0.3.2-AW-2024-08-30
input  0 dim 3 640 640 1, data_format=2, quant_format=0, name=input/output[0], none-quant
output 0 dim 80 80 64 1, data_format=0, name=uid_17_out_0b_uid_1_out_0, none-quant
output 1 dim 80 80 1 1, data_format=0, name=uid_16_out_0b_uid_1_out_0, none-quant
output 2 dim 80 80 51 1, data_format=0, name=uid_15_out_0b_uid_1_out_0, none-quant
output 3 dim 40 40 64 1, data_format=0, name=uid_14_out_0b_uid_1_out_0, none-quant
output 4 dim 40 40 1 1, data_format=0, name=uid_13_out_0b_uid_1_out_0, none-quant
output 5 dim 40 40 51 1, data_format=0, name=uid_12_out_0b_uid_1_out_0, none-quant
output 6 dim 20 20 64 1, data_format=0, name=uid_11_out_0b_uid_1_out_0, none-quant
output 7 dim 20 20 1 1, data_format=0, name=uid_10_out_0b_uid_1_out_0, none-quant
output 8 dim 20 20 51 1, data_format=0, name=uid_9_out_0ub_uid_1_out_0, none-quant
nbg name=model/yolo11s-pose_9_uint8_a733.nb, size: 7284048.
create network 0: 16110 us.
prepare network: 3977 us.
buffer ptr: 0x202f5380, buffer size: 1228800
network: 0, loop count: 1
run time for this network 0: 32374 us.
output 0, ptr 0x20421480, size 409600.
output 1, ptr 0x205b1500, size 6400.
output 2, ptr 0x205b7980, size 326400.
output 3, ptr 0x206f6640, size 102400.
output 4, ptr 0x2075a6c0, size 1600.
output 5, ptr 0x2075c040, size 81600.
output 6, ptr 0x207abbc0, size 25600.
output 7, ptr 0x207c4c80, size 400.
output 8, ptr 0x207c5340, size 20400.
post process time : 4 ms
detection num: 3
 0:  94%, [ 370,    0,  589,  346], person
405.75 26.20 = 0.96988
419.11 23.03 = 0.96501
405.65 21.63 = 0.29929
441.04 31.18 = 0.99146
421.11 22.33 = 0.04379
455.76 67.51 = 0.99977
430.35 62.14 = 0.99950
466.39 121.18 = 0.99797
405.08 109.99 = 0.98330
447.50 96.32 = 0.98985
382.14 70.42 = 0.94582
466.06 166.69 = 0.99986
455.44 165.19 = 0.99974
411.43 242.60 = 0.99939
497.02 230.87 = 0.99880
408.66 307.99 = 0.98213
562.98 301.11 = 0.97806
 0:  88%, [  86,   27,  292,  389], person
146.77 66.48 = 0.99659
157.56 60.56 = 0.99517
138.54 62.15 = 0.93738
177.10 58.15 = 0.97191
136.75 58.00 = 0.22714
182.16 88.95 = 0.99876
146.17 100.58 = 0.99755
210.99 144.46 = 0.99757
161.48 152.14 = 0.98186
171.08 179.70 = 0.99453
131.07 188.60 = 0.97326
222.98 197.17 = 0.99975
178.42 204.61 = 0.99950
250.05 264.09 = 0.99831
151.41 290.25 = 0.99650
287.21 296.16 = 0.97016
127.74 355.67 = 0.95755
 0:  92%, [ 228,   39,  399,  407], person
275.86 94.61 = 0.99351
286.44 88.28 = 0.98999
267.42 87.73 = 0.88035
308.03 73.30 = 0.97833
265.48 74.83 = 0.23741
339.54 98.91 = 0.99963
280.47 109.74 = 0.99938
372.16 125.90 = 0.99505
272.82 170.12 = 0.98157
380.93 163.22 = 0.98073
243.21 204.51 = 0.94730
339.07 225.60 = 0.99986
302.82 223.45 = 0.99980
294.02 310.02 = 0.99952
314.44 286.69 = 0.99926
270.90 355.43 = 0.99344
374.00 318.30 = 0.99277
destroy npu finished.
~NpuUint.

This performance data only calculates the time consumption of model inference. Unless otherwise specified, it does not include the time consumption of pre-processing and post-processing.

SoCNPUModelInput ResolutionNetwork Creation TimeNetwork Preparation TimeSingle Frame Inference TimePost-processing TimeTotal TimeFrame Rate
Allwinner A733Vivante VIP9000yolo11s-pose640×64016.1 ms4.0 ms32.4 ms4.0 ms56.5 ms30.9 FPS

    You need to be logged into GitHub to post a comment. If you are already logged in, please ignore this message.

    Radxa-docs © 2026 by Radxa Computer (Shenzhen) Co.,Ltd. is licensed under CC BY 4.0