AVTok: 1D Unified Tokenization for Holistic Audio-Video Generation

Audio-video generation has recently gained unprecedented research attention, aiming to synthesize high-quality sounding video content with fine-grained synchronization and semantic alignment between the auditory and visual components. The preceding methods predominantly adopt a dual-branch design with separate tokenization and generation modules per modality, neglecting the representation gap while necessitating intensive computational resources for proper training. Inspired by recent advancements in one-dimensional visual tokenization, we present AVTok, a novel unified tokenizer designated for holistic audio-video generation. AVTok features a dual-stream transformer-based architecture with shared encoder-decoder and modal-specific learnable queries to efficiently and effectively encode an audio-video pair into a compact one-dimensional latent representation with a unified codebook. To cope with the heterogeneous information imbalance that hinders AVTok from exploiting aligned audio-visual information, we devise a hierarchical training strategy to progressively realize reconstruction capabilities for each modality. Extensive experiments demonstrate that AVTok excels both in audio-video reconstruction and when integrated into downstream pipelines for audio-to-video, video-to-audio, and class-conditional joint audio-video generation. AVTok paves the way for the challenge of joint audio-video tokenization and provides a potential direction to build unified large multimodal models for audio-video generation.
Read Original

Related

Papers with Code paper 3d ago

WanSong v1.0 Technical Report

Music generation foundation models have recently attracted significant industry attention. However, achieving efficient generation and high-fidelity long-form audio while supportin...