Search papers, labs, and topics across Lattice.
University of Hong Kong https, github, MCG-NJU/UniDDT
2
0
3
UniDDT achieves a groundbreaking balance between multimodal understanding and generation, outperforming existing models in both tasks with enhanced semantic coherence.
High-dimensional discrete tokens, previously out of reach for generative models, can now be directly generated, unlocking a unified token prediction paradigm for multimodal architectures.
