This paper analyzes the feasibility of position-assisted ultra-wideband multicast audio and derives bounds on one-listener arrival spread and all-receiver block delivery.

This paper provides a comprehensive review of audio editing in the era of foundation models, including a unified taxonomy of existing editing tasks and major foundation-model paradigms.

The paper introduces AudioWeave, a unified model for text-to-audio and audio editing tasks.

The authors present StreamMUSE, an inference system for language models that performs synchronous generation with external signals in real-time.

This paper introduces Diff2Mix, a generative automatic music mixing system using diffusion models and a differentiable mixing console for stylistic control.

The paper proposes a probabilistic off-tree forwarding mechanism to reduce latency in large-scale distributed systems.

The paper introduces TORUS, a self-coherence test for evaluating the agreement between audio understanding, generation, and editing capabilities in unified audio models.

This paper presents a causal, training-free linker that improves the performance of streaming end-to-end diarizers by tracking more speakers than their output slots.

This paper proposes two learned approaches to reuse pretrained perceptual knowledge for multichannel spatial audio.

This paper introduces Speaker Anonymization (SA) as a novel perturbation mechanism for zero-shot voice conversion, balancing timbre leakage and prosodic utility while enabling strictly causal, zero-lo…

The paper presents ScenA, a method for generating rich conversational audio using a text-to-audio flow-matching foundation model conditioned on multiple reference voices and a free-form natural langua…

This paper proposes a three-stage generative framework for speech bandwidth extension that achieves state-of-the-art results with one-step inference.

This paper proposes a deep learning-assisted multicast subgrouping framework to improve content delivery in massive MIMO multicasting.