Commit History
8 days ago
-
Add Core AI depth decoder (fused depth_frame) for FluidAudio TTS (FluidAudio #49)
Verified signatureTTS on the Core AI backend runs the fused depth_frame (3.24 ms/frame, codes 99.40% identical to HF). FluidAudio TTS RTFx 4.28 -> 5.07, round-trip WER 2.5%. Request: https://github.com/FluidInference/FluidAudio/issues/49#issuecomment-65016582636dc374f
