Audio 2 Face microservice

Description

Convert audio to blendshape files and stream them.

This microservice contains an audio processing pipeline, receiving audio file as input, sending audio (wav) and blendshape files (json) as output.

Usage

This microservice needs parameters:

• streamNumber: maximum number of streams that the pipeline will use.

Supported Platforms

The current supported architecture is x86.

Performance

Can support 25 streams at 30 FPS on a machine with a RTX 3090 + 20 cpu cores

Deployment requirements

• Gpu: RTX 3080+

License

Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. NVIDIA Corporation and its licensors retain all intellectual property and proprietary rights in and to this software, related documentation and any modifications thereto. Any use, reproduction, disclosure or distribution of this software and related documentation without an express license agreement from NVIDIA Corporation is strictly prohibited.

Known Issues / Limitations

Currently, resampling is not supported, 16 kHz audio wav files as input are required.