Shypd.aiThe largest open index of AI toolsRAD-NeRF is an open-source video generation tool that synthesizes real-time neural talking portraits via audio-spatial decomposition. It allows for the creation of realistic and synchronized facial animations from audio input.
Trending
About
RAD-NeRF is an open-source PyTorch re-implementation of the paper "Real-time Neural Radiance Talking Portrait Synthesis via Audio-spatial Decomposition." This tool enables the creation of highly realistic and synchronized talking portraits from audio input, effectively generating deepfakes and audio-driven avatars. It leverages advanced neural radiance fields to achieve real-time performance. The repository provides detailed instructions for installation, data pre-processing, and usage, including training head and torso models, finetuning lips, and running inference with or without a graphical user interface. It supports both DeepSpeech and Wav2Vec for audio feature extraction and offers options for custom backgrounds.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
