AI Paper Atlas

Research map / Label-efficient and robust learning / Noisy labels

GOMPSNR: Reflourish the Signal-to-Noise Ratio Metric for Audio Generation Tasks

Lingling Dai, Andong Li, Cheng Chi, Yifan Liang, Xiaodong Li, Chengshi Zheng

AAAI 2026 · Label-efficient and robust learning › Noisy labels · approach: Contrastive learning on noisy data

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
replace IP with omnidirectional phase derivatives in SNR (OMPSNR) (ablation)SNR · LibriTTS, pretrained VocosPCC/SRCC with perceptual metrics: SNR PCC/SRCC not exceeding 0.1 -> significantly higherImproved
rectify correlation component (GOMPSNR vs OMPSNR) (ablation)OMPSNR · LibriTTS, pretrained VocosPCC/SRCC with perceptual metricsImproved
add vanilla phase loss P (ablation)Vocos · LJSpeechPESQ: 3.749 -> 3.711No clear effect
add OP loss (ablation)Vocos · LJSpeechPESQ: 3.749 -> 3.752No clear effect
weight OP loss by magnitude (WOP) (ablation)Vocos · LJSpeechPESQ: 3.752 -> 3.928Improved
add vanilla RI loss on top of WOP (ablation)Vocos + WOP · LJSpeechPESQ: 3.928 -> 3.891Hurt
replace RI with OmniRI (L1) (ablation)Vocos + WOP · LJSpeechPESQ: 3.891 -> 3.998Improved
use CORI instead of ORI (L1) (ablation)Vocos + WOP · LJSpeechPESQ: 3.998 -> 3.992No clear effect
use L2 instead of L1 distance in ORI (ablation)Vocos + WOP + ORI · LJSpeechPESQ: 3.998 -> 3.791Hurt
linear magnitude loss instead of log (ablation)APNet · LJSpeechPESQ / GOMPSNR: PESQ 2.836 -> 3.124; GOMPSNR 3.335 -> -0.957Mixed
linear magnitude loss instead of log (ablation)Vocos · LibriTTSPESQ: 3.167 -> 3.298Improved
full proposed loss combination (Lin, WOP, CORI)APNet2 · LibriTTSPESQ: 1.685 -> 3.789Improved

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