This is a Speaker Recognition project adapted from Speaker Recognition to work with ROS.
- SciPy
- scikit-learn
- scikits.talkbox
- pyssp
- PyQt
- PyAudio
- (Optional)bob.
Run
cd my_speaker
make -C gmm/ to compile the fast gmm implementation. Require gcc >= 4.7.
It will be used as default, if successfully compiled.
You should understand that real-time speaker recognition is extremely hard, because we only use corpus of about 1 second length to identify the speaker. Therefore the real-time system doesn't work very perfect.
usage: speaker-recognition.py [-h] -t TASK -i INPUT -m MODEL
Speaker Recognition Command Line Tool
optional arguments:
-h, --help show this help message and exit
-t TASK, --task TASK Task to do. Either "enroll" or "predict"
-i INPUT, --input INPUT
Input Files(to predict) or Directories(to enroll)
-m MODEL, --model MODEL
Model file to save(in enroll) or use(in predict)
Wav files in each input directory will be labeled as the basename of the directory.
Note that wildcard inputs should be *quoted*, and they will be sent to glob module.
Examples:
Train:
./speaker-recognition.py -t enroll -i "./new_data/train/*" -m model.out
Test:
./speaker-recognition.py -t predict -i "./new_data/test/*.wav" -m model.out
ROS
-there are 2 Nodes in this package.
1.record :
Continuously record voice and publish the audio as a numpy array on the topic /wav
2.predict:
subscribes to the topic /wav and if voice activity is detected it tries to predict the speaker.
publishes on the topic /speaker the speaker it predicts
running ROS
source ./ros_ws/devel/setup.bash
roslaunch speaker speaker_reco.launch