Adaptive approach for sparse representations using the locally competitive algorithm for audio , available at https://github.com/SoufiyanBAHADI/ALCA. This is the code for the preprint available at https://doi.org/10.1109/MLSP52302.2021.9596348
The required python packages are listed in requirements.txt, and can be installed with:
pip install -r requirements.txt
cd ALCA
python main.py
optional arguments:
-h, --help shows this help message and exit
-p PATH, --path PATH The path of the data set.
--tau TAU Neurons' time constant.
--dt DT Euler's resolution method clock.
--threshold THRESHOLD
Firing threshold.
--stride STRIDE Stride size.
--ker-len KER_LEN Kernels' length.
--num-chan NUM_CHAN Number of channels.
--iters ITERS The LCA's iterations.
--optimizer {sgd,adam}
The optimizer needed for training.
--lr LR Learning rate.
--batch-size BATCH_SIZE
the size of each mini batch.
--buffer-size BUFFER_SIZE
The size of the buffer where to store steady states
for backpropagation through time algorithm
-e EPOCHS, --epochs EPOCHS
number of epochs.
--eval Specifies the evaluation. If false the algorithm will
run in training mode
-v, --verbose allows the program verbosity
--random-init parameters are initiallized randomly
--resume RESUME The epoch from which the learning will resume
--plot If specified the program will plot all outputs. --eval
should be specified
© Copyright (June 2021) Soufiyan Bahadi, prof. Jean Rouat, prof. Éric Plourde. University of Sherbrooke. NEuro COmputational & Intelligent Signal Processing Research Group (NECOTIS)