What are some good open source projects exploring emotional voice synthesis?

There are tons of TTS software out there, but they don't incorporate human emotions during speech synthesis. For example, anger, tiredness, surprise, happiness...

What solutions exist for this today?