Memorization of melodies by complex-valued recurrent network
Makoto Kinouchi, Masafumi Hagiwara · 2002
Memorization of the temporal sequences is very important and interesting problem. One of the examples is memorization of melodies. When we memorize melodies, we can recall it from a part. The function can be considered as an associative memory of the temporal sequences. From a technical point of view, it can be considered that the rest of the melody can be determined by some leading some notes. In this paper, we apply the complex-valued recurrent network, which has much superior ability to deal with temporal sequences than the conventional real-valued networks, to memorization of melodies. It is shown in computer simulations that the network can memorize some melodies and recall them correctly from any part.