https://github.com/ZtoYtoQ/NCNN_Android_SinglePoseEstimation

NCNN Android SinglePoseEstimation Demo
https://github.com/ZtoYtoQ/NCNN_Android_SinglePoseEstimation

Last synced: 2 months ago
JSON representation

NCNN Android SinglePoseEstimation Demo

• Host: GitHub
• URL: https://github.com/ZtoYtoQ/NCNN_Android_SinglePoseEstimation
• Owner: ZtoYtoQ
• License: mit
• Created: 2020-04-26T07:01:07.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-04-26T08:26:35.000Z (over 4 years ago)
• Last Synced: 2024-08-01T05:13:16.676Z (5 months ago)
• Language: Java
• Size: 5.99 MB
• Stars: 14
• Watchers: 3
• Forks: 3
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-ncnn - NCNN_Android_SinglePoseEstimation

README

        
# NCNN_Android_SinglePoseEstimation

## Introduction 

The simple project mainly provides Single Pose Estimation demo in Android though NCNN .

## Usage

### S1:

1. Download ncnn-android-vulkan-lib.zip for [NCNN Releases](https://github.com/Tencent/ncnn/releases)

### S2:

```bash

git clonehttps://github.com/ZtoYtoQ/NCNN_Android_SinglePoseEstimation.git

unzip ncnn-android-vulkan-lib.zip  { clone root-dir}/app/src/main/jni 

```

### S3:

Open this project with Android Studio, build it and run.

**NOTE**: this project with  SDK version: 26 , Gradle model version=5.6.4 NDK version: 20.0.5594570

## Running Time

|               **Device**               | Thread : 1 | Threads : 2 |

| :------------------------------------: | :--------: | :---------: |

|      Mi 10 ( Snapdragon 865 CPU )      |  ~640 ms   |   ~350 ms   |

| Samsung S7 edge ( Snapdragon 820 CPU ) |  ~1400 ms  |   ~900 ms   |

## Demo

![demo_result](https://github.com/ZtoYtoQ/NCNN_Android_SinglePoseEstimation/blob/master/demo_result.jpg)

## Reference

NCNN-android-image classification :   https://github.com/nihui/ncnn-android-squeezenet

## License

[MIT](https://choosealicense.com/licenses/mit/)