Python based, ENF extraction from video/audio recordings.
Python Version Tested = 3.6.13 Environments used are mentioned in requirements.txt
Python = 3.9 is currently being tested with the libraries updated to their current version.
- In the previous version 3.6.13, the scipy library usage was restricted to <= 1.2.0 since the function 'imresize' was used to resize array dimensions.
- The dependency is removed by replacing the function with Pillow library based resize function.
- The updated pyenf module works for audio recordings. The video recordings are still being tested.
To test if the required python libraries are installed properly for audio recording based ENF estimation, run the following code with the input recording. A sample power recording is available in Rec_demo folder.
python pyenf.py --file INPUT_FILENAME --nominal NOMINAL_FREQUENCY
The defaults are already set to Rec_demo/power_recording.wav file and 60 Hz nominal frequency value. So, by simply running the program without any input should be enough to verify code functionality
python pyenf.py