Reduced-Order Modeling of Deep Neural Networks
This addresses efficiency issues for users of deep neural networks, but it appears incremental as it builds on existing reduced-order modeling techniques.
The paper tackles the problem of speeding up deep neural network inference by introducing a method that replaces convolutional layers with smaller fully-connected layers, achieving a relatively small drop in accuracy in many practical cases.
We introduce a new method for speeding up the inference of deep neural networks. It is somewhat inspired by the reduced-order modeling techniques for dynamical systems.The cornerstone of the proposed method is the maximum volume algorithm. We demonstrate efficiency on neural networks pre-trained on different datasets. We show that in many practical cases it is possible to replace convolutional layers with much smaller fully-connected layers with a relatively small drop in accuracy.