AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

SLD-L2S: Hierarchical Subspace Latent Diffusion for High-Fidelity Lip to Speech Synthesis

Yifan Liang, Andong Li, Kang Yang, Guochen Yu, Fangkun Liu, Lingling Dai, Xiaodong Li, Chengshi Zheng

AAAI 2026 · Image generation › Autoregressive and latent generation · approach: Diffusion audio synthesis · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use DiCB backbone instead of parameter-matched DiT (ablation)SLD-L2S · LRS3-TEDUTMOS: 4.0835 -> 4.2096Improved
use DiCB backbone instead of DiT (ablation)SLD-L2S · LRS3-TEDSECS: 0.618 -> 0.804Improved
use reparameterized flow matching instead of velocity prediction (ablation)SLD-L2S · LRS3-TEDUTMOS: 3.9540 -> 4.2096Improved
use reparameterized flow matching instead of velocity prediction (ablation)SLD-L2S · LRS3-TEDWER: 29.70 -> 30.22No clear effect
add SLM loss (ablation)SLD-L2S · LRS3-TEDUTMOS: 4.2885 -> 4.2096Hurt
add SLM loss (ablation)SLD-L2S · LRS3-TEDD-BERT and SECS: D-BERT 0.8207 -> 0.8284; SECS 0.776 -> 0.804Improved
add semantic loss (ablation)SLD-L2S · LRS3-TEDUTMOS: 4.1859 -> 4.2096Mixed
use 8 subspaces instead of 32 (ablation)SLD-L2S · LRS3-TEDUTMOS: 4.1288 -> 4.2096Improved
use 8 subspaces instead of 4 (ablation)SLD-L2S · LRS3-TEDUTMOS: 4.1797 -> 4.2096Improved
vary DiCB kernel size (3x5 to 9x11) (ablation)SLD-L2S · LRS3-TEDUTMOS: 4.2120 / 4.1914 / 4.2008 vs 4.2096No clear effect
SLD-L2S vs V2SFlow in subjective intelligibilitySLD-L2S · LRS3 MOSintelligibility MOS: 3.69 (V2SFLow) vs 3.65Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy