FreeSR
FreeSR copied to clipboard
nonday
A Free Library for Speaker Recognition (Verification)，implemented by ncnn.
FreeSR: A Free Library for Speaker Recognition
This is a free library for Speaker Recognition (Verification), which is completely implemented in C/C++ by ncnn. In addition, Speaker Gender Recognition is also implemented. We also provide a example for java.
FreeSR
| Windows (X64/Release) | Linux (X64) | Android | |
|---|---|---|---|
| >=VS2015 | g++7.5.0 | AS 3.2 | |
| Speaker Recognition(Verification) | ✔ | ✔ | ✔ |
| Speaker Gender Recognition | ✔ | ✔ | ✔ |
Usage
mkdir build; cd build; cmake.exe .. -G"Visual Studio 14 Win64"
The input is preferably raw data without normalization from wav (single channel, bitspersample 16bits and sample rate 16K). More detailed information about usage of FreeSR can be referred to speaker_recognizer.h.
Example
Performance
| Platform | Input Length | Time(ms) | Library Size | 1:N |
|---|---|---|---|---|
| Win10 i7-8700 CPU @ 3.20GHz | 4s@16KHz | -- | 5.22MB(dll) | 1:100000 |
nonday