In this study, we investigate leveraging cross-attention control for efficient audio editing within auto-regressive models. Inspired by image editing methodologies, we develop a Prompt-to-Prompt-like ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results