| CPC G06N 3/04 (2013.01) [G06F 3/165 (2013.01); G10L 21/10 (2013.01)] | 16 Claims |
|
1. An electronic apparatus, comprising:
a memory configured to store a trained first artificial intelligence model; and
a processor connected to the memory and configured to:
based on receiving an input audio signal, divide the input audio signal into a plurality of sections having a predetermined length, obtain a plurality of input frequency spectrum images representing a plurality of frequency spectrums corresponding to the plurality of sections,
input the plurality of input frequency spectrum images to the trained first artificial intelligence model,
obtain a plurality of output frequency spectrum images from the trained first artificial intelligence model,
obtain a final output image by stitching the plurality of output frequency spectrum images, and
obtain an output audio signal based on the final output image,
wherein the first artificial intelligence model is trained based on a target learning image, and
wherein the target learning image represents a target frequency spectrum of a specific style, and is obtained from a trained second artificial intelligence model based on a random value and a condition value corresponding to the specific style, and
wherein the second artificial intelligence model is trained to obtain a plurality of target learning images representing a plurality of target frequency spectrums of a plurality of styles based on a plurality of condition values corresponding to the plurality of styles.
|