LatentFlowSR achieves superior audio super-resolution by generating high-resolution latents from low-resolution ones via conditional flow matching in a noise-robust autoencoder latent space.
Audio super resolution using neural networks, In Kuleshov, Volodymyr and Enam, S Zayd and Ermon, Stefano.Proc
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LatentFlowSR: High-Fidelity Audio Super-Resolution via Noise-Robust Latent Flow Matching
LatentFlowSR achieves superior audio super-resolution by generating high-resolution latents from low-resolution ones via conditional flow matching in a noise-robust autoencoder latent space.