EmoRES-TTS Audio Demo

Mixed-emotion speech examples comparing the native model, conventional CoCoEmo steering, and residual-enhanced steering.

For the clearest comparison, use headphones and keep playback volume fixed across systems.

Speaker prompts and ground-truth recordings cannot be displayed due to dataset licensing restrictions.

01 sad 0.67, angry 0.33

“Then help me stay here mom.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
02 sad 0.67, angry 0.33

“Things have never turned out anything the way that I thought they would be.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
03 sad 0.50, angry 0.25

“Why is that?”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
04 sad 0.67, angry 0.33

“And I didn't want to take any of it.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
05 sad 0.67, angry 0.33

“I'm trying the best I can.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
06 angry 0.67, sad 0.33

“You know- you know how this works. I have to go.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
07 angry 0.83, surprise 0.17

“That's all you can offer me? I mean, my the stuff in there is worth way more than fifty dollar.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
08 surprise 0.67, angry 0.33

“What?”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3
09 happy 0.67, surprise 0.33

“Oh good idea.”

No-steer
Native conditioning only
CoCoEmo
λc = 1, λr = 1
EmoRES (Ours)
λc = 1, λr = 3