# DeRA-MOS: Optimizing Text-to-Music Evaluation via Decoupled Listwise Ranking and Modality Alignment

arXiv: https://arxiv.org/abs/2606.10010 · HTML: https://mlsanity.com/paper/2606.10010/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 26 of its 40 references that are in the map.
- First posted on arXiv 8 Jun 2026.

## Builds on (most cited first, 26 of 26)

- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Gradient Surgery for Multi-Task Learning](https://mlsanity.com/paper/2001.06782/) (2020)
- [HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis](https://mlsanity.com/paper/2010.05646/) (2020)
- [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://mlsanity.com/paper/2211.06687/) (2022)
- [AudioLDM: Text-to-Audio Generation with Latent Diffusion Models](https://mlsanity.com/paper/2301.12503/) (2023)
- [Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech](https://mlsanity.com/paper/2106.06103/) (2021)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [MusicLM: Generating Music From Text](https://mlsanity.com/paper/2301.11325/) (2023)
- [UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022](https://mlsanity.com/paper/2204.02152/) (2022)
- [AudioCLIP: Extending CLIP to Image, Text and Audio](https://mlsanity.com/paper/2106.13043/) (2021)
- [Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models](https://mlsanity.com/paper/2301.12661/) (2023)
- [NISQA: A Deep CNN-Self-Attention Model for Multidimensional Speech Quality Prediction with Crowdsourced Datasets](https://mlsanity.com/paper/2104.09494/) (2021)
- [Rank consistent ordinal regression for neural networks with application to age estimation](https://mlsanity.com/paper/1901.07884/) (2019)
- [MOSNet: Deep Learning based Objective Assessment for Voice Conversion](https://mlsanity.com/paper/1904.08352/) (2019)
- [Label Distribution Learning](https://mlsanity.com/paper/1408.6027/) (2014)
- [Adapting Frechet Audio Distance for Generative Music Evaluation](https://mlsanity.com/paper/2311.01616/) (2023)
- [Generalization Ability of MOS Prediction Networks](https://mlsanity.com/paper/2110.02635/) (2021)
- [MuQ: Self-Supervised Music Representation Learning with Mel Residual Vector Quantization](https://mlsanity.com/paper/2501.01108/) (2025)
- [MBNet: MOS Prediction for Synthesized Speech with Mean-Bias Network](https://mlsanity.com/paper/2103.00110/) (2021)
- [LDNet: Unified Listener Dependent Modeling in MOS Prediction for Synthetic Speech](https://mlsanity.com/paper/2110.09103/) (2021)
- [MusicEval: A Generative Music Dataset with Expert Ratings for Automatic Text-to-Music Evaluation](https://mlsanity.com/paper/2501.10811/) (2025)
- [The AudioMOS Challenge 2025](https://mlsanity.com/paper/2509.01336/) (2025)
- [QAMRO: Quality-aware Adaptive Margin Ranking Optimization for Human-aligned Assessment of Audio Generation Systems](https://mlsanity.com/paper/2508.08957/) (2025)
- [DRASP: A Dual-Resolution Attentive Statistics Pooling Framework for Automatic MOS Prediction](https://mlsanity.com/paper/2508.21407/) (2025)
- [ASTAR-NTU solution to AudioMOS Challenge 2025 Track1](https://mlsanity.com/paper/2507.09904/) (2025)

## Abstract (arXiv)

Evaluating text-to-music (TTM) systems remains expensive because music impression (MI) and text alignment (TA) scores rely on human mean opinion scores (MOS). Most automatic MOS estimators are trained with point-wise regression or distributional classification. These objectives do not directly optimize rank-based metrics and provide weak geometric constraints for cross-modal coherence. To address these gaps, we propose DeRA-MOS, a decoupled optimization framework for TTM evaluation. For MI, we introduce a batch-aware listwise ranking loss that models relative order within each mini-batch and better aligns with evaluation based on Spearman's rank correlation coefficient (SRCC). For TA, we introduce a score-anchored modality alignment loss that maps human scores to target audio-text similarity and regularizes the latent space before fusion. By effectively mitigating the point-wise training mismatch and modality drift, experiments on MusicEval demonstrate that our decoupled framework yields substantial improvements in both MI and TA ranking metrics, establishing a robust paradigm for large-scale TTM evaluation.