libfacedetection-master
所属分类：其他
开发工具：C/C++
文件大小：13544KB
下载次数：3
上传日期：2020-02-10 13:59:25
上 传 者：
att0206
说明： 这是一个 项目它主要实现的功能是对人脸关键点的检测
(this is a file that I want to download a file)
文件列表:
CMakeLists.txt (2524, 2019-03-18)
ChangeLog (1611, 2019-03-18)
LICENSE (1803, 2019-03-18)
aarch64-toolchain.cmake (321, 2019-03-18)
example (0, 2019-03-18)
example\libfacedetect.cpp (5613, 2019-03-18)
example\libfacedetectcnn-example.cpp (3622, 2019-03-18)
images (0, 2019-03-18)
images\chloecalmon.png (790085, 2019-03-18)
images\cnnresult.png (779789, 2019-03-18)
images\keliamoniz1.jpg (64424, 2019-03-18)
images\keliamoniz2.jpg (88213, 2019-03-18)
models (0, 2019-03-18)
models\caffe (0, 2019-03-18)
models\caffe\yufacedetectnet-open-v1.caffemodel (2703121, 2019-03-18)
models\caffe\yufacedetectnet-open-v1.prototxt (17480, 2019-03-18)
models\caffe\yufacedetectnet-open-v1.solver.prototxt (352, 2019-03-18)
models\caffe\yufacedetectnet-open-v1.train.prototxt (19464, 2019-03-18)
models\openvino (0, 2019-03-18)
models\openvino\yufacedetectnet-open-v1-320x240.bin (2693440, 2019-03-18)
models\openvino\yufacedetectnet-open-v1-320x240.xml (42208, 2019-03-18)
src (0, 2019-03-18)
src\facedetectcnn-floatdata.cpp (15636456, 2019-03-18)
src\facedetectcnn-int8data.cpp (2893278, 2019-03-18)
src\facedetectcnn-model.cpp (24868, 2019-03-18)
src\facedetectcnn.cpp (43790, 2019-03-18)
src\facedetectcnn.h (15079, 2019-03-18)
# libfacedetection
This is an open source library for CNN-based face detection in images. The CNN model has been converted to static variables in C source files. The source code does not depend on any other libraries. What you need is just a C++ compiler. You can compile the source code under Windows, Linux, ARM and any platform with a C++ compiler.
SIMD instructions are used to speed up the detection. You can enable AVX2 if you use Intel CPU or NEON for ARM.
The model file has also been provided in directory ./models/.
examples/libfacedetectcnn-example.cpp shows how to use the library.
## How to Compile
* Please add -O3 to turn on optimizations when you compile the source code using g++.
* Please choose 'Maximize Speed/-O2' when you compile the source code using Microsoft Visual Studio.
Create a build folder:
```
mkdir build; cd build; rm -rf *
```
### Cross build for aarch***
1. set cross compiler for aarch*** (please refer to aarch***-toolchain.cmake)
2. set opencv path since the example code depends on opencv
```
cmake \
-DENABLE_INT8=ON \
-DENABLE_NEON=ON \
-DCMAKE_BUILD_TYPE=RELEASE \
-DCMAKE_TOOLCHAIN_FILE=../aarch***-toolchain.cmake \
..
make
```
### Native build for avx2
```
cmake \
-DENABLE_INT8=ON \
-DENABLE_AVX2=ON \
-DCMAKE_BUILD_TYPE=RELEASE \
-DDEMO=ON \
..
make
```
## CNN-based Face Detection on Windows
| Method |Time | FPS |Time | FPS |
|--------------------|--------------|-------------|--------------|-------------|
| | X*** |X*** | X*** |X*** |
| |Single-thread |Single-thread|Multi-thread |Multi-thread |
|OpenCV Haar+AdaBoost (***0x480)| -- | -- | 12.33ms | 81.1 |
|cnn (CPU, ***0x480) | ***.21ms | 15.57 | 15.59ms | ***.16 |
|cnn (CPU, 320x240) | 15.23ms | 65.68 | 3.99ms | 250.40 |
|cnn (CPU, 160x120) | 3.47ms | 288.08 | 0.95ms | 1052.20 |
|cnn (CPU, 128x96) | 2.35ms | 425.95 | 0.***ms | 1562.10 |
* OpenCV Haar+AdaBoost runs with minimal face size 48x48
* Face detection only, and no landmark detection included.
* Minimal face size ~12x12
* Intel(R) Core(TM) i7-7700 CPU @ 3.6GHz.
## CNN-based Face Detection on ARM Linux (Raspberry Pi 3 B+)
| Method |Time | FPS |Time | FPS |
|--------------------|--------------|-------------|--------------|-------------|
| |Single-thread |Single-thread|Multi-thread |Multi-thread |
|cnn (CPU, ***0x480) | 512.04ms | 1.95 | 174.89ms | 5.72 |
|cnn (CPU, 320x240) | 123.47ms | 8.10 | 42.13ms | 23.74 |
|cnn (CPU, 160x120) | 27.42ms | 3***7 | 9.75ms | 102.58 |
|cnn (CPU, 128x96) | 17.78ms | 56.24 | 6.12ms | 163.50 |
* Face detection only, and no landmark detection included.
* Minimal face size ~12x12
* Raspberry Pi 3 B+, Broadcom BCM2837B0, Cortex-A53 (ARMv8) ***-bit SoC @ 1.4GHz
## Author
* Shiqi Yu,
## Contributors
* Jia Wu
* Shengyin Wu
* Dong Xu
## Acknowledgment
The work is partly supported by the Science Foundation of Shenzhen (Grant No. JCYJ20150324141711699).
近期下载者：
相关文件：
收藏者：