Bias values of the model

[keunwoochoi]

import h5py
model_name = "vggnet5"
keras_filename = "vggnet5_local_keras_model_CNN_stft_11_frame_173_freq_257_folding_0_best.keras"

f = h5py.File(keras_filename)

In auralise.py, I only use param_0, which is weights. Bias values are in param_1

>>> f['layer_3']['param_0']
<HDF5 dataset "param_0": shape (64, 64, 3, 3), type "<f4">
>>> f['layer_3']['param_1']
<HDF5 dataset "param_1": shape (64,), type "<f4">

[aseemw]

Hi Keunwoo,

How have you learned these CNN weights? Which dataset and task have you used? '

Thanks,
Aseem

[keunwoochoi]

I used a private dataset, some details are described in the paper - experiment section.