← all emotions
Contemplation
Best adapter: rank 16 , checkpoint seen 3336 · EmoNet-z -0.56 (baseline 0.06) · blend 0.635 · genu 0.171 · speaker-sim 0.248.
Instruction prompt: GENERAL: A voice that is adult; extremely carrying meditation; 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 -1.44 ±0.01 · bl 0.2 · gn 0.1 ▲ high -1.43
▼ low -1.45
z mean -1.24 ±0.41 · bl 0.6 · gn 0.1 ▲ high -0.41
▼ low -1.45
z mean -1.43 ±0.03 · bl 0.7 · gn 0.2 ▲ high -1.36
▼ low -1.44
with emo reference
z mean -1.42 ±0.04 · bl 0.2 · gn 0.5 ▲ high -1.35
▼ low -1.45
z mean -0.51 ±0.48 · bl 0.0 · gn 0.4 ▲ high -0.07
▼ low -1.43
z mean -0.35 ±0.59 · bl 0.2 · gn 0.5 ▲ high +0.09
▼ low -1.43
Sentence 2
no reference
z mean -0.86 ±0.75 · bl 0.6 · gn 0.1 ▲ high +0.38
▼ low -1.45
z mean -1.43 ±0.00 · bl 0.5 · gn 0.1 ▲ high -1.43
▼ low -1.44
z mean -1.34 ±0.17 · bl 0.7 · gn 0.2 ▲ high -1.00
▼ low -1.45
with emo reference
z mean -0.32 ±0.48 · bl 0.5 · gn 0.4 ▲ high +0.24
▼ low -1.20
z mean -0.93 ±0.63 · bl 0.1 · gn 0.2 ▲ high +0.01
▼ low -1.44
z mean -0.76 ±0.61 · bl 0.2 · gn 0.6 ▲ high -0.02
▼ low -1.44
Sentence 3
no reference
z mean -0.78 ±0.37 · bl 0.1 · gn 0.1 ▲ high -0.31
▼ low -1.43
z mean +0.00 ±0.48 · bl 0.3 · gn 0.1 ▲ high +0.47
▼ low -0.91
z mean +0.05 ±0.28 · bl 0.7 · gn 0.2 ▲ high +0.51
▼ low -0.29
with emo reference
z mean -0.50 ±0.08 · bl 0.1 · gn 0.3 ▲ high -0.41
▼ low -0.64
z mean -0.24 ±0.63 · bl 0.0 · gn 0.1 ▲ high +0.41
▼ low -1.43
z mean -0.87 ±0.43 · bl 0.2 · gn 0.6 ▲ high -0.36
▼ low -1.43
Sentence 4
no reference
z mean -1.42 ±0.04 · bl 0.2 · gn 0.1 ▲ high -1.33
▼ low -1.44
z mean -1.23 ±0.46 · bl 0.5 · gn 0.1 ▲ high -0.30
▼ low -1.47
z mean -0.84 ±0.69 · bl 0.6 · gn 0.2 ▲ high +0.31
▼ low -1.45
with emo reference
z mean -0.54 ±0.49 · bl 0.2 · gn 0.4 ▲ high +0.06
▼ low -1.36
z mean +0.11 ±0.23 · bl 0.0 · gn 0.5 ▲ high +0.38
▼ low -0.31
z mean -0.19 ±0.09 · bl 0.3 · gn 0.6 ▲ high -0.05
▼ low -0.32
Sentence 5
no reference
z mean -1.43 ±0.00 · bl 0.2 · gn 0.1 ▲ high -1.43
▼ low -1.43
z mean -1.27 ±0.35 · bl 0.6 · gn 0.1 ▲ high -0.57
▼ low -1.45
z mean -0.75 ±0.42 · bl 0.6 · gn 0.1 ▲ high -0.26
▼ low -1.40
with emo reference
z mean -0.97 ±0.53 · bl 0.2 · gn 0.3 ▲ high -0.28
▼ low -1.44
z mean -0.85 ±0.73 · bl 0.0 · gn 0.4 ▲ high +0.43
▼ low -1.45
z mean -0.69 ±0.45 · bl 0.1 · gn 0.3 ▲ high -0.17
▼ low -1.43