基于多模态流匹配的少样本声学生成
Few-shot Acoustic Synthesis with Multimodal Flow Matching
把流匹配 + 扩散 Transformer 用于以稀疏多模态场景线索（空间、几何、声学）为条件的概率化生成，其条件融合与少样本泛化思路可迁移到视觉中的跨模态条件生成任务；AGREE 这种联合嵌入做几何一致性检索评测的做法，也可借鉴到视觉生成结果的分布级评测。
原文依据 FLAC leverages a diffusion transformer trained with a flow-matching objective to generate RIRs at arbitrary positions in novel scenes
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论生成与场景声学一致的音频对于沉浸式虚拟环境至关重要。近期的神经声场方法能够实现空间连续的声音渲染，但仍局限于特定场景，需要密集的音频测量并为每个环境付出高昂的训练成本。少样本方法提升了跨房间的可扩展性，但仍依赖多次录音，且由于是确定性的，无法在稀疏上下文下刻画场景声学固有的不确定性。我们提出 FLow-matching ACoustic generation（FLAC），一种用于少样本声学生成的概率方法，它在给定极少场景上下文的条件下对合理房间冲激响应（RIR）的分布进行建模。FLAC 利用以流匹配目标训练的扩散 Transformer，在空间、几何与声学线索的条件下，为新场景中任意位置生成 RIR。在 AcousticRooms 和 Hearing Anything Anywhere 数据集上，FLAC 以单样本超越了最优的八样本基线。为补充标准感知指标，我们进一步提出 AGREE，一种联合声学-几何嵌入，可通过检索与分布度量对生成的 RIR 进行几何一致性评估。本工作是首次将生成式流匹配应用于显式 RIR 合成，为稳健且数据高效的声学生成开辟了新方向。项目主页：https://amandinebtto.github.io/FLAC/
英文原文摘要 1455 字符
Generating audio that is acoustically consistent with a scene is essential for immersive virtual environments. Recent neural acoustic field methods enable spatially continuous sound rendering but remain scene-specific, requiring dense audio measurements and costly training for each environment. Few-shot approaches improve scalability across rooms but still rely on multiple recordings and, being deterministic, fail to capture the inherent uncertainty of scene acoustics under sparse context. We introduce FLow-matching ACoustic generation (FLAC), a probabilistic method for few-shot acoustic synthesis that models the distribution of plausible room impulse responses (RIRs) given minimal scene context. FLAC leverages a diffusion transformer trained with a flow-matching objective to generate RIRs at arbitrary positions in novel scenes, conditioned on spatial, geometric, and acoustic cues. FLAC outperforms state-of-the-art eight-shot baselines with one-shot on both the AcousticRooms and Hearing Anything Anywhere datasets. To complement standard perceptual metrics, we further introduce AGREE, a joint Acoustic-GeometRy EmbEdding, enabling geometry-consistent evaluation of generated RIRs through retrieval and distributional metrics. This work is the first to apply generative flow matching to explicit RIR synthesis, establishing a new direction for robust and data-efficient acoustic synthesis. Project page: https://amandinebtto.github.io/FLAC/