用3×3卷积核+极少量数据，就能自动生成音乐效果。
CAK: Emergent Audio Effects from Minimal Deep Learning
- 用一个可学习的卷积核，结合软门控机制实现控制参数平滑调节。
- 仅用200个样本训练，就生成了基于输入特征的频率依赖时移效果。
- 适合音频效果设计、小样本生成任务的研究者和创作者。
我们展示了一个3×3卷积核在仅200个个性化样本上训练后，能产生涌现的音频效果。通过两种关键技术实现：(1) 条件感知核（CAK），输出 = 输入 + (学习模式 × 控制值)，并采用软门控机制在控制值为零时保持身份不变；(2) 审计生成对抗网络（AuGAN），将对抗训练从“真假判断”转为“是否应用指定控制值”。网络不再学习生成或检测伪造，而是协作验证控制参数的执行，从而发现独特变换。学习到的卷积核呈现对角结构，引发频率相关的时序偏移，能够根据输入特征生成音乐化效果。结果表明，对抗训练可在极小数据下发现音频变换，为效果设计开辟新路径。
原文摘要 · Abstract (English)
We demonstrate that a single 3x3 convolutional kernel can produce emergent audio effects when trained on 200 samples from a personalized corpus. We achieve this through two key techniques: (1) Conditioning Aware Kernels (CAK), where output = input + (learned_pattern x control), with a soft-gate mechanism supporting identity preservation at zero control; and (2) AuGAN (Audit GAN), which reframes adversarial training from "is this real?" to "did you apply the requested value?" Rather than learning to generate or detect forgeries, our networks cooperate to verify control application, discovering unique transformations. The learned kernel exhibits a diagonal structure creating frequency-dependent temporal shifts that are capable of producing musical effects based on input characteristics. Our results show the potential of adversarial training to discover audio transformations from minimal data, enabling new approaches to effect design.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。