arXiv：cs.LG· Wanli Sun, Anton Ragni·· 4 小时前
面向基于能量 TTS 模型的 Score-Based 训练方法
Score-Based Training for Energy-Based TTS Models
AI 导读
针对基于能量的模型（EBM）训练，该论文提出一种新的训练准则，学习更适合一阶优化方案的 score，以改进 NCE 和 sliced score matching（SSM）忽略对数似然函数形式的问题。实验对比了这些方法在 EBM 训练上的表现，论文已被 Interspeech 2025 接收。
正文
Abstract:Noise contrastive estimation (NCE) is a popular method for training energy-based models (EBM) with intractable normalisation terms. The key idea of NCE is to learn by comparing unnormalised log-likelihoods of the reference and noisy samples, thus avoiding explicitly computing normalisation terms. However, NCE critically relies on the quality of noisy samples. Recently, sliced score matching (SSM) has been popularised by closely related diffusion models (DM). Unlike NCE, SSM learns a gradient of log-likelihood, or score, by learning distribution of its projections on randomly chosen directions. However, both NCE and SSM disregard the form of log-likelihood function, which is problematic given that EBMs and DMs make use of first-order optimisation during inference. This paper proposes a new criterion that learns scores more suitable for first-order schemes. Experiments contrasts these approaches for training EBMs.
| Comments: | 5 pages, 3 figures, 4 tables. Accepted to Interspeech 2025 |
| Subjects: | Sound (cs.SD); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS) |
| Cite as: | arXiv:2505.13771 [cs.SD] |
| (or arXiv:2505.13771v2 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2505.13771 arXiv-issued DOI via DataCite |
Submission history
From: Wanli Sun [view email]
[v1]
Mon, 19 May 2025 23:12:25 UTC (68 KB)
[v2]
Wed, 14 Oct 2026 07:23:09 UTC (66 KB)
来源：arXiv：cs.LG · arxiv.org