MAG: Multi-Modal Aligned Autoregressive Co-Speech Gesture Generation without Vector Quantization

SLR
Re-assign community