Learning of physical-like sound synthesis models by adaptive spline recurrent neural networks

A recently introduced neural networks architecture, ´adaptive spline neural networks´ with FIR/IIR synapse, is used to define a general class of physical-like sound synthesis model. To reduce computational cost, use is made of power-of-two synapses followed by a CR-spline-based flexible activation function the shape of which can be modified through its control points. The learning phase is performed by an efficient combinatorial optimisation algorithm, Tabu Search, for both power-of-two weights and CR-spline control points