System and method for synthetic audio generation

The system comprising:

• an input/output interface to receive a textual description with one or more details to generate an audio file, wherein the textual description includes a type of sound, a duration of sound required, bitrate of the sound, and a sampling frequency of the audio file to be generated;
• one or more hardware processors;
• a memory in communication with the one or more hardware processors, wherein the one or more hardware processors are configured to execute programmed instructions stored in the memory, to:
• pre-process the received textual description to extract keywords using a natural language processing (NLP) technique;
• determine a contextual relationship among the extracted keywords and corresponding the received textual description to identify a category of the sound in the audio file, wherein the category of the sound includes a base sound and one or more background noises;
• select at least one model corresponding to the identified base sound and the one or more background noises from a pre-created model repository, wherein the selected at least one model of the pre-created model repository is developed based on a generative adversarial network;
• and
• generate the audio files using the selected at least one model, wherein the generated audio file is of the predefined sampling frequency, bitrate, and duration.

A non-transitory computer readable medium storing one or more instructions which when executed by one or more processors on a system, cause the one or more processors to perform method comprising:

• receiving, via an input/output interface, a textual description with one or more details to generate an audio file, wherein the textual description includes a type of sound, a duration of sound required, bitrate of the sound, and a sampling frequency of the audio file to be generated;
• pre-processing, via one or more hardware processors, the received textual description to extract keywords using a natural language processing (NLP) technique;
• determining, via the one or more hardware processors, a contextual relationship among the extracted keywords and corresponding the received textual description to identify a category of the sound in the audio file, wherein the category of the sound includes a base sound and one or more background noises;
• selecting, via the one or more hardware processors, at least one model corresponding to the identified base sound and the one or more background noises from a pre-created model repository, wherein the selected at least one model of the pre-created model repository is developed based on a generative adversarial network;
• and
• generating, via the one or more hardware processors, the audio files using the selected at least one model, wherein the generated audio file is of the predefined sampling frequency, bitrate, and duration.