DCT-Conv: Coding filters in convolutional networks with Discrete Cosine Transform
This addresses efficiency and overtraining issues in deep learning for practitioners, though it is incremental as it builds on prior work using fewer trained parameters.
The paper tackles the problem of convolutional neural networks being data-greedy and slow to train by coding filters with trained DCT parameters, showing improved performance over traditional convolution and maintaining good results even when 99.9% of parameters are switched off.
Convolutional neural networks are based on a huge number of trained weights. Consequently, they are often data-greedy, sensitive to overtraining, and learn slowly. We follow the line of research in which filters of convolutional neural layers are determined on the basis of a smaller number of trained parameters. In this paper, the trained parameters define a frequency spectrum which is transformed into convolutional filters with Inverse Discrete Cosine Transform (IDCT, the same is applied in decompression from JPEG). We analyze how switching off selected components of the spectra, thereby reducing the number of trained weights of the network, affects its performance. Our experiments show that coding the filters with trained DCT parameters leads to improvement over traditional convolution. Also, the performance of the networks modified this way decreases very slowly with the increasing extent of switching off these parameters. In some experiments, a good performance is observed when even 99.9% of these parameters are switched off.