ResGen predicts cumulative vector embeddings of masked RVQ tokens, decoupling generative sampling cost from token depth and improving FID and TTS metrics over autoregressive baselines.
Training details A.1
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Efficient Generative Modeling with Residual Vector Quantization-Based Tokens
ResGen predicts cumulative vector embeddings of masked RVQ tokens, decoupling generative sampling cost from token depth and improving FID and TTS metrics over autoregressive baselines.