LivePortrait

LivePortrait 是一种基于深度学习的人像驱动生成模型，可将静态人像转换为具有真实表情与动作的动态视频效果。

手动下载模型并上传到 raspberrypi5，或者通过以下命令拉取模型仓库。

提示

如果没有安装 git lfs，先参考git lfs 安装说明进行安装。

git clone https://huggingface.co/AXERA-TECH/LivePortrait

文件说明：

m5stack@raspberrypi:~/rsp/LivePortrait $ ls -lh
total 16K
drwxrwxr-x 3 m5stack m5stack 4.0K Aug 13 09:41 assets
-rw-rw-r-- 1 m5stack m5stack    0 Aug 13 09:41 config.json
drwxrwxr-x 5 m5stack m5stack 4.0K Aug 13 09:41 python
-rw-rw-r-- 1 m5stack m5stack 5.7K Aug 13 09:41 README.md

创建虚拟环境

python -m venv lvpr

激活虚拟环境

source lvpr/bin/activate

安装依赖包

pip install https://github.com/AXERA-TECH/pyaxengine/releases/download/0.1.3.rc2/axengine-0.1.3-py3-none-any.whl
pip install -r python/requirements.txt
pip install requests tqdm

Image

运行：

python ./python/infer.py --source ./assets/examples/source/s0.jpg --driving ./assets/examples/driving/d8.jpg --models ./python/axmodels/ --output-dir ./axmodel_infer

运行结果：

(lvpr) m5stack@raspberrypi:~/rsp/LivePortrait $ python ./python/infer.py --source ./assets/examples/source/s0.jpg --driving ./assets/examples/driving/d8.jpg --models ./python/axmodels/ --output-dir ./axmodel_infer
[INFO] Available providers:  ['AXCLRTExecutionProvider']
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 144960ad
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 144960ad
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 0f7260e8
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 144960ad
FaceAnalysisDIY warmup time: 0.237s
LandmarkRunner warmup time: 0.183s
2025-08-13 10:04:42.914 | INFO     | __main__:main:727 - Start making driving motion template...
2025-08-13 10:04:43.782 | INFO     | __main__:main:747 - Prepared pasteback mask done.
2025-08-13 10:04:44.565 | INFO     | __main__:main:787 - The output of image-driven portrait animation is an image.
2025-08-13 10:04:50.566 | DEBUG    | __main__:warp_decode:647 - warp time: 5.997s
2025-08-13 10:04:50.902 | INFO     | __main__:main:881 - Animated image: ./axmodel_infer/s0--d8.jpg
2025-08-13 10:04:50.902 | INFO     | __main__:main:882 - Animated image with concat: ./axmodel_infer/s0--d8_concat.jpg
2025-08-13 10:04:50.920 | DEBUG    | __main__:<module>:894 - LivePortrait axmodel infer time: 18.068s 

原图：

输出：

Video

运行：

python ./python/infer.py --source ./assets/examples/source/s0.jpg --driving ./assets/examples/driving/d0.mp4 --models ./python/axmodels/ --output-dir ./axmodel_infer

运行结果：

(lvpr) m5stack@raspberrypi:~/rsp/LivePortrait $ python ./python/infer.py --source ./assets/examples/source/s0.jpg --driving ./assets/examples/driving/d0.mp4 --models ./python/axmodels/ --output-dir ./axmodel_infer

[INFO] Available providers:  ['AXCLRTExecutionProvider']
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 144960ad
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 144960ad
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 0f7260e8
[INFO] Using provider: AXCLRTExecutionProvider
[INFO] SOC Name: AX650N
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Compiler version: 3.3 144960ad
FaceAnalysisDIY warmup time: 0.213s
LandmarkRunner warmup time: 0.180s
2025-08-13 10:14:47.779 | INFO     | __main__:main:727 - Start making driving motion template...
2025-08-13 10:15:04.110 | INFO     | __main__:main:747 - Prepared pasteback mask done.
2025-08-13 10:15:04.903 | INFO     | __main__:main:785 - The animated video consists of 78 frames.
2025-08-13 10:15:11.468 | DEBUG    | __main__:warp_decode:647 - warp time: 6.561s
2025-08-13 10:25:50.630 | DEBUG    | __main__:warp_decode:647 - warp time: 8.343s
2025-08-13 10:25:51.493 | INFO     | __main__:has_audio_stream:114 - Error occurred while probing video: ./assets/examples/driving/d0.mp4, you may need to install ffprobe! (https://ffmpeg.org/download.html) Now set audio to false!
2025-08-13 10:25:54.608 | INFO     | __main__:main:870 - Animated video: ./axmodel_infer/s0--d0.mp4
2025-08-13 10:25:54.609 | INFO     | __main__:main:871 - Animated video with concat: ./axmodel_infer/s0--d0_concat.mp4
2025-08-13 10:25:54.644 | DEBUG    | __main__:<module>:894 - LivePortrait axmodel infer time: 670.930s 

原视频：

输出：

Next 目录索引

Linux PC

CM4Stack

CoreMP135

AI 加速卡

LLM-8850 Card

快速上手

视觉模型

大语言模型

多模态模型

音频模型

生成模型

应用列表

进阶使用

大语言模型

实时 AI 语音助手

OpenAI 语音助手

小智语音助手

小聆语音助手

火山引擎语音助手

离线语音识别

Unit ASR

工业控制

StamPLC

IoT 测量仪表

Air Quality

PowerHub

Module13.2 PPS

VAMeter

T-Lite

Ezdata

Ethernet 摄像头

PoECAM

Wi-Fi 摄像头

TimerCAM

Unit CamS3/-5MP

AI 摄像头

UnitV2

StickV/UnitV

LoRa & LoRaWAN

TTN (The Things Network)

Meshtastic

电机驱动

Unit Roller485/CAN

开发工具

爱好套装