Skip to main content
Fig. 12 | EURASIP Journal on Audio, Speech, and Music Processing

Fig. 12

From: Variational Autoencoders for chord sequence generation conditioned on Western harmonic music complexity

Fig. 12

Analysis of participants’ negative ratings: among users who disagree with the complexity value indicated by our models, we evaluate how many of them consider that complexity has a different value from the one indicated in the test. In the first example (a), most users express that the complexity is higher. In the second (b), about half indicates a higher value, while the other half assigns a lower complexity. The case described in b is defined as “ambiguous” sample

Back to article page