Table 1 Basic music theory reward
From: MetaMGC: a music generation framework for concerts in metaverse
Condition | Meaning | Reward | |
|---|---|---|---|
\(R^{m1}(s_{t:1}, a_{t})\) | \(a_{t-1}=a_{t},(if\ a_{t-3} = a_{t-2}= a_{t-1})\) | Pitch recurrence 4 times | 0.9 |
\(a_{t-1} \ne a_{t},(if\ a_{t} \ne a_{t})\) | Pitch recurrence 3 times | 0 | |
\(R^{m2}(s_{t:1}, a_{t})\) | \(\mid a_{t}-a_{t-1}\mid \le 12\) | Intervals greater than an octave | 0.1 |
\(\mid a_{t}-a_{t-1} \mid >12\) | Intervals less than an octave | − 0.8 | |
\(R^{m3}(s_{t:1}, a_{t})\) | \(a_{t} \in [ a_{min}, a_{max} ]\) | Within the set vocal range | 0.1 |
\(a_{t} \notin [ a_{min}, a_{max} ]\) | Not in the set vocal range | − 0.8 |