LAV: Audio-Driven Dynamic Visual Generation with Neural Compression and StyleGAN2
This work addresses the challenge of audio-driven visual generation for artistic and computational applications, presenting an incremental improvement by combining existing models in a novel way.
The paper tackles the problem of generating dynamic visual content from audio by integrating EnCodec's neural audio compression with StyleGAN2, resulting in a system that produces nuanced and semantically coherent audio-visual translations without explicit feature mappings.
This paper introduces LAV (Latent Audio-Visual), a system that integrates EnCodec's neural audio compression with StyleGAN2's generative capabilities to produce visually dynamic outputs driven by pre-recorded audio. Unlike previous works that rely on explicit feature mappings, LAV uses EnCodec embeddings as latent representations, directly transformed into StyleGAN2's style latent space via randomly initialized linear mapping. This approach preserves semantic richness in the transformation, enabling nuanced and semantically coherent audio-visual translations. The framework demonstrates the potential of using pretrained audio compression models for artistic and computational applications.