AI 中文总结
针对现有音频效果建模依赖干参考的问题，提出无干参考的对比学习框架RelFx，通过双分支暹罗编码器等实现相对效果表征，在Fx风格迁移任务上优于现有方法
AI 中文摘要
音频效果（Fx）表征学习在智能音乐制作中发挥关键作用，包括自动混音和Fx风格迁移。现有方法通常依赖干信号或近干信号参考进行效果建模，但实际中很少有真正未处理的音频，因为真实录音不可避免地会反映麦克风、房间声学及先前的信号处理。我们认为，与其追求绝对效果编码，音频信号间的相对效果距离对实际音乐制作更有意义。受此启发，我们提出RelFx，这是一种对比学习框架，可从通用音频集合中学习相对效果变换，表征训练期间无需干参考。我们的方法采用配备交叉注意力和差分门控融合的双分支暹罗编码器，从参考片段和经效果处理的、内容相关的片段中推断共享效果变换。我们进一步提出一种用于双向效果编码的反对称融合变体，即交换输入顺序可直接产生近符号反转的嵌入，这是早期工作未探索的特性。此外，我们的无干参考公式消除了对干多轨数据集的依赖，可在含效果的音频上进行训练。在Fx风格迁移实验中，我们在标准Fx-Encoder++ MUSDB18评估协议下展现出最先进性能，在所有四类乐器类别中均持续优于现有方法。
英文摘要
Audio effects (Fx) representation learning plays a key role in intelligent music production, including automatic mixing and Fx style transfer. Existing methods typically rely on dry or nearly dry references for effect modeling, yet truly unprocessed audio is rarely available in practice, as real recordings inevitably reflect the microphone, room acoustics, and preceding signal processing. Instead of pursuing absolute effect encodings, we argue that the relative effect distance between audio signals is more meaningful for real-world music production. Motivated by this, we propose RelFx, a contrastive learning framework that learns relative effect transformations from general audio collections without requiring dry references during representation training. Our approach uses a dual-branch Siamese encoder equipped with cross-attention and differential gating fusion to infer the shared effect transformation from a reference clip and an effect-processed, content-related clip. We further propose an antisymmetric fusion variant for bidirectional effect encoding, such that swapping the input order directly produces a nearly sign-reversed embedding, a property not explored in earlier work. Moreover, our dry-reference-free formulation eliminates the reliance on dry multitrack datasets and enables training on effect-bearing audio. Experiments on Fx style transfer demonstrate state-of-the-art performance under the standard Fx-Encoder++ MUSDB18 evaluation protocol, consistently outperforming existing approaches across all four instrument categories.
Comments8 pages (6 pages of main text), 3 figures, 3 tables. Accepted at the 27th International Society for Music Information Retrieval Conference (ISMIR 2026). Project page: https://relative-fx.github.io
