**1. Introduction**

Advances in technology and computer science have greatly enhanced the possibility of designing, developing, and deploying intelligent musical devices. A typical well-studied subset of these intelligent devices are IoMusTs (Internet of Musical Things). According to [1], an IoMusT is a "computing device capable of sensing and exchanging data to serve a musical purpose." An IoMusT does not need to be able to produce, select, or modify music, but it can be any device that is "music aware" in the sense that its behavior is directly related to music. As an example, PixMob devices [2] have been widely used in musical performances. These devices that can be either worn (smartband), thrown (balls), or attached to the audience seats and are able to produce light patterns synchronized with life performances.

Not all intelligent musical devices are IoMusts. We can design intelligent devices where the intelligence is embedded in the device, and thus, we may say that we gave an Intelligent Musical dEvice (IME) but not as a part of the Internet of Musical Things. In [3], the evolution of the design of intelligent musical instruments is studied. In most cases, these instruments use artificial intelligence as a tool for user interaction without requiring any connection to public networks or cloud-based services. It is important to consider that machine learning ML, in most of these cases, cannot be considered as an independent agent but mainly as one of the possible alternatives for designing layers of a complete system. These types of devices can also be considered as cyber-physical systems, as they clearly require intelligent software systems and dedicated hardware.

In this work, we will create a framework that covers all, or at least a very wide part, of intelligent musical devices and helps design, understand, and study them.

The rest of the paper is divided as follows: First, in the Materials and Methods section, the taxonomy is and published works are detailed, as well as the analysis methodology used to test the different systems. The results obtained for the different systems are then detailed and explained in the Results and Discussion sections. Finally, conclusions are presented.
