← all emotions
Interest
Best adapter: rank 64 , checkpoint seen 1672 · EmoNet-z -0.59 (baseline -1.039) · blend 0.421 · genu 0.129 · speaker-sim 0.288.
Instruction prompt: GENERAL: A voice that is adult, with extremely carrying fascination and genuine in delivery.
SCRIPT:
Five neutral sentences, each in English / Deutsch / Français, no reference vs with emotion reference (a same-emotion training clip supplies the voice; the prompt applies the emotion).
Each cell = 5 seeds. We show mean±std of the emotion score and play the highest and lowest seed (top = strongest emotion, bottom = weakest), since a fixed seed pins the sampling and single clips vary.
Sentence 1
no reference
z mean -1.01 ±0.13 · bl 0.1 · gn 0.0 ▲ high -0.88
▼ low -1.24
z mean -0.97 ±0.32 · bl 0.3 · gn 0.1 ▲ high -0.58
▼ low -1.54
z mean -1.27 ±0.28 · bl 0.5 · gn 0.1 ▲ high -0.93
▼ low -1.76
with emo reference
z mean -0.94 ±0.03 · bl 0.0 · gn 0.3 ▲ high -0.89
▼ low -0.97
z mean -0.45 ±0.13 · bl 0.0 · gn 0.2 ▲ high -0.25
▼ low -0.60
z mean -1.00 ±0.02 · bl 0.1 · gn 0.2 ▲ high -0.96
▼ low -1.02
Sentence 2
no reference
z mean -0.34 ±0.13 · bl 0.7 · gn 0.2 ▲ high -0.08
▼ low -0.47
z mean +0.11 ±0.15 · bl 0.4 · gn 0.1 ▲ high +0.35
▼ low -0.09
z mean +0.45 ±0.22 · bl 0.5 · gn 0.2 ▲ high +0.67
▼ low +0.03
with emo reference
z mean -0.37 ±0.16 · bl 0.3 · gn 0.3 ▲ high -0.18
▼ low -0.58
z mean -0.05 ±0.16 · bl 0.1 · gn 0.1 ▲ high +0.22
▼ low -0.24
z mean -0.27 ±0.10 · bl 0.1 · gn 0.2 ▲ high -0.09
▼ low -0.39
Sentence 3
no reference
z mean -0.81 ±0.13 · bl 0.2 · gn 0.1 ▲ high -0.65
▼ low -1.03
z mean -0.28 ±0.28 · bl 0.4 · gn 0.1 ▲ high -0.04
▼ low -0.74
z mean -0.56 ±0.27 · bl 0.6 · gn 0.2 ▲ high -0.29
▼ low -0.93
with emo reference
z mean -0.62 ±0.15 · bl 0.1 · gn 0.1 ▲ high -0.44
▼ low -0.78
z mean -0.58 ±0.21 · bl 0.1 · gn 0.1 ▲ high -0.38
▼ low -0.96
z mean -0.19 ±0.10 · bl 0.1 · gn 0.1 ▲ high -0.10
▼ low -0.38
Sentence 4
no reference
z mean -0.34 ±0.21 · bl 0.1 · gn 0.1 ▲ high +0.01
▼ low -0.63
z mean +0.15 ±0.32 · bl 0.4 · gn 0.1 ▲ high +0.66
▼ low -0.35
z mean +0.72 ±0.18 · bl 0.4 · gn 0.2 ▲ high +0.89
▼ low +0.41
with emo reference
z mean -0.28 ±0.11 · bl 0.1 · gn 0.2 ▲ high -0.11
▼ low -0.41
z mean -0.22 ±0.14 · bl 0.0 · gn 0.1 ▲ high +0.07
▼ low -0.32
z mean +0.09 ±0.21 · bl 0.2 · gn 0.3 ▲ high +0.28
▼ low -0.30
Sentence 5
no reference
z mean -0.93 ±0.10 · bl 0.1 · gn 0.0 ▲ high -0.82
▼ low -1.06
z mean -0.71 ±0.22 · bl 0.4 · gn 0.1 ▲ high -0.45
▼ low -1.03
z mean -0.35 ±0.42 · bl 0.5 · gn 0.1 ▲ high +0.30
▼ low -0.73
with emo reference
z mean -0.32 ±0.19 · bl 0.1 · gn 0.2 ▲ high -0.02
▼ low -0.58
z mean -0.57 ±0.13 · bl 0.0 · gn 0.1 ▲ high -0.38
▼ low -0.68
z mean -0.56 ±0.13 · bl 0.1 · gn 0.2 ▲ high -0.41
▼ low -0.74