Audio enhancement demo

Original vs. CleanUNet + FlashSR speech samples

Compare paired recordings from Spanish and Catalan speech datasets. Each row keeps the transcript, source audio, and enhanced version together so noise reduction and upsampling differences are easier to hear.