Pushing the Frontier of Audiovisual Perception with Large-Scale Multimodal Correspondence Learning

We introduce Perception Encoder Audiovisual, PE-AV, a new family of encoders for audio and video understanding trained with scaled contrastive learning. Built on PE, PE-AV makes several key contributions to extend representations to audio, and natively support joint embeddings across audio-video, audio-text, and video-text modalities. PE-AV's unified cross-modal embeddings enable novel tasks such as speech retrieval, and set a new state of the art across standard audio and video benchmarks. We unlock this by building a strong audiovisual data engine that synthesizes high-quality captions for O(100M) audio-video pairs, enabling large-scale supervision consistent across modalities. Our audio data includes speech, music, and general sound effects-avoiding single-domain limitations common in prior work. We exploit ten pairwise contrastive objectives, showing that scaling cross-modality and caption-type pairs strengthens alignment and improves zero-shot performance. We further develop PE-A-Frame by fine-tuning PE-AV with frame-level contrastive objectives, enabling fine-grained audio-frame-to-text alignment for tasks such as sound event detection.

MauBERT: Universal Phonetic Inductive Biases for Few-Shot Acoustic Units Discovery

This paper introduces MauBERT, a multilingual extension of HuBERT that leverages articulatory features for robust cross-lingual phonetic representation learning. We continue HuBERT pre-training with supervision based on a phonetic-to-articulatory feature mapping in 55 languages. Our models learn from multilingual data to predict articulatory features or phones, resulting in language-independent representations that capture multilingual phonetic properties. Through comprehensive ABX discriminability testing, we show MauBERT models produce more context-invariant representations than state-of-the-art multilingual self-supervised learning models. Additionally, the models effectively adapt to unseen languages and casual speech with minimal self-supervised fine-tuning (10 hours of speech). This establishes an effective approach for instilling linguistic inductive biases in self-supervised speech models.

Real-Time Streamable Generative Speech Restoration with Flow Matching

Diffusion-based generative models have greatly impacted the speech processing field in recent years, exhibiting high speech naturalness and spawning a new research direction. Their application in real-time communication is, however, still lagging behind due to their computation-heavy nature involving multiple calls of large DNNs. Here, we present Stream.FM, a frame-causal flow-based generative model with an algorithmic latency of 32 milliseconds (ms) and a total latency of 48 ms, paving the way for generative speech processing in real-time communication. We propose a buffered streaming inference scheme and an optimized DNN architecture, show how learned few-step numerical solvers can boost output quality at a fixed compute budget, explore model weight compression to find favorable points along a compute/quality tradeoff, and contribute a model variant with 24 ms total latency for the speech enhancement task. Our work looks beyond theoretical latencies, showing that high-quality streaming generative speech processing can be realized on consumer GPUs available today. Stream.FM can solve a variety of speech processing tasks in a streaming fashion: speech enhancement, dereverberation, codec post-filtering, bandwidth extension, STFT phase retrieval, and Mel vocoding. As we verify through comprehensive evaluations and a MUSHRA listening test, Stream.FM establishes a state-of-the-art for generative streaming speech restoration, exhibits only a reasonable reduction in quality compared to a non-streaming variant, and outperforms our recent work (Diffusion Buffer) on generative streaming speech enhancement while operating at a lower latency.

Kunnafonidilaw ka Cadeau: an ASR dataset of present-day Bambara

We present Kunkado, a 160-hour Bambara ASR dataset compiled from Malian radio archives to capture present-day spontaneous speech across a wide range of topics. It includes code-switching, disfluencies, background noise, and overlapping speakers that practical ASR systems encounter in real-world use. We finetuned Parakeet-based models on a 33.47-hour human-reviewed subset and apply pragmatic transcript normalization to reduce variability in number formatting, tags, and code-switching annotations. Evaluated on two real-world test sets, finetuning with Kunkado reduces WER from 44.47\% to 37.12\% on one and from 36.07\% to 32.33\% on the other. In human evaluation, the resulting model also outperforms a comparable system with the same architecture trained on 98 hours of cleaner, less realistic speech. We release the data and models to support robust ASR for predominantly oral languages.

Enhancing Fully Formatted End-to-End Speech Recognition with Knowledge Distillation via Multi-Codebook Vector Quantization

Conventional automatic speech recognition (ASR) models typically produce outputs as normalized texts lacking punctuation and capitalization, necessitating post-processing models to enhance readability. This approach, however, introduces additional complexity and latency due to the cascaded system design. In response to this challenge, there is a growing trend to develop end-to-end (E2E) ASR models capable of directly predicting punctuation and capitalization, though this area remains underexplored. In this paper, we propose an enhanced fully formatted E2E ASR model that leverages knowledge distillation (KD) through multi-codebook vector quantization (MVQ). Experimental results demonstrate that our model significantly outperforms previous works in word error rate (WER) both with and without punctuation and capitalization, and in punctuation error rate (PER). Evaluations on the LibriSpeech-PC test-clean and test-other subsets show that our model achieves state-of-the-art results.

In-Context Audio Control of Video Diffusion Transformers

Recent advancements in video generation have seen a shift towards unified, transformer-based foundation models that can handle multiple conditional inputs in-context. However, these models have primarily focused on modalities like text, images, and depth maps, while strictly time-synchronous signals like audio have been underexplored. This paper introduces In-Context Audio Control of video diffusion transformers (ICAC), a framework that investigates the integration of audio signals for speech-driven video generation within a unified full-attention architecture, akin to FullDiT. We systematically explore three distinct mechanisms for injecting audio conditions: standard cross-attention, 2D self-attention, and unified 3D self-attention. Our findings reveal that while 3D attention offers the highest potential for capturing spatio-temporal audio-visual correlations, it presents significant training challenges. To overcome this, we propose a Masked 3D Attention mechanism that constrains the attention pattern to enforce temporal alignment, enabling stable training and superior performance. Our experiments demonstrate that this approach achieves strong lip synchronization and video quality, conditioned on an audio stream and reference images.

Reliable Audio Deepfake Detection in Variable Conditions via Quantum-Kernel SVMs

Detecting synthetic speech is challenging when labeled data are scarce and recording conditions vary. Existing end-to-end deep models often overfit or fail to generalize, and while kernel methods can remain competitive, their performance heavily depends on the chosen kernel. Here, we show that using a quantum kernel in audio deepfake detection reduces falsepositive rates without increasing model size. Quantum feature maps embed data into high-dimensional Hilbert spaces, enabling the use of expressive similarity measures and compact classifiers. Building on this motivation, we compare quantum-kernel SVMs (QSVMs) with classical SVMs using identical mel-spectrogram preprocessing and stratified 5-fold cross-validation across four corpora (ASVspoof 2019 LA, ASVspoof 5 (2024), ADD23, and an In-the-Wild set). QSVMs achieve consistently lower equalerror rates (EER): 0.183 vs. 0.299 on ASVspoof 5 (2024), 0.081 vs. 0.188 on ADD23, 0.346 vs. 0.399 on ASVspoof 2019, and 0.355 vs. 0.413 In-the-Wild. At the EER operating point (where FPR equals FNR), these correspond to absolute false-positiverate reductions of 0.116 (38.8%), 0.107 (56.9%), 0.053 (13.3%), and 0.058 (14.0%), respectively. We also report how consistent the results are across cross-validation folds and margin-based measures of class separation, using identical settings for both models. The only modification is the kernel; the features and SVM remain unchanged, no additional trainable parameters are introduced, and the quantum kernel is computed on a conventional computer.

Smark: A Watermark for Text-to-Speech Diffusion Models via Discrete Wavelet Transform

Text-to-Speech (TTS) diffusion models generate high-quality speech, which raises challenges for the model intellectual property protection and speech tracing for legal use. Audio watermarking is a promising solution. However, due to the structural differences among various TTS diffusion models, existing watermarking methods are often designed for a specific model and degrade audio quality, which limits their practical applicability. To address this dilemma, this paper proposes a universal watermarking scheme for TTS diffusion models, termed Smark. This is achieved by designing a lightweight watermark embedding framework that operates in the common reverse diffusion paradigm shared by all TTS diffusion models. To mitigate the impact on audio quality, Smark utilizes the discrete wavelet transform (DWT) to embed watermarks into the relatively stable low-frequency regions of the audio, which ensures seamless watermark-audio integration and is resistant to removal during the reverse diffusion process. Extensive experiments are conducted to evaluate the audio quality and watermark performance in various simulated real-world attack scenarios. The experimental results show that Smark achieves superior performance in both audio quality and watermark extraction accuracy.

Task Vector in TTS: Toward Emotionally Expressive Dialectal Speech Synthesis

Recent advances in text-to-speech (TTS) have yielded remarkable improvements in naturalness and intelligibility. Building on these achievements, research has increasingly shifted toward enhancing the expressiveness of generated speech, such as dialectal and emotional TTS. However, cross-style synthesis combining both dialect and emotion remains challenging and largely unexplored, mainly due to the scarcity of dialectal data with emotional labels. To address this, we propose Hierarchical Expressive Vector (HE-Vector), a two-stage method for Emotional Dialectal TTS. In the first stage, we construct different task vectors to model dialectal and emotional styles independently, and then enhance single-style synthesis by adjusting their weights, a method we refer to as Expressive Vector (E-Vector). For the second stage, we hierarchically integrate these vectors to achieve controllable emotionally expressive dialect synthesis without requiring jointly labeled data, corresponding to Hierarchical Expressive Vector (HE-Vector). Experimental results demonstrate that HE-Vectors achieve superior performance in dialect synthesis, and promising results in synthesizing emotionally expressive dialectal speech in a zero-shot setting.

Asynchronous Pipeline Parallelism for Real-Time Multilingual Lip Synchronization in Video Communication Systems

This paper introduces a parallel and asynchronous Transformer framework designed for efficient and accurate multilingual lip synchronization in real-time video conferencing systems. The proposed architecture integrates translation, speech processing, and lip-synchronization modules within a pipeline-parallel design that enables concurrent module execution through message-queue-based decoupling, reducing end-to-end latency by up to 3.1 times compared to sequential approaches. To enhance computational efficiency and throughput, the inference workflow of each module is optimized through low-level graph compilation, mixed-precision quantization, and hardware-accelerated kernel fusion. These optimizations provide substantial gains in efficiency while preserving model accuracy and visual quality. In addition, a context-adaptive silence-detection component segments the input speech stream at semantically coherent boundaries, improving translation consistency and temporal alignment across languages. Experimental results demonstrate that the proposed parallel architecture outperforms conventional sequential pipelines in processing speed, synchronization stability, and resource utilization. The modular, message-oriented design makes this work applicable to resource-constrained IoT communication scenarios including telemedicine, multilingual kiosks, and remote assistance systems. Overall, this work advances the development of low-latency, resource-efficient multimodal communication frameworks for next-generation AIoT systems.