arXiv subjects

Reinhold Haeb-Umbach

Publications and source records attributed to Reinhold Haeb-Umbach.

At least 19 recordsLinked to original sources

arXiv2026

Investigating the Integration of Spatial Information in Foundation-Model-Based Speaker Diarization

Spatial information gleaned from multi-channel input has been shown to lead to improvements in meeting processing tasks like diarization and source separation. At the same time, diarization based on features extracted by large pretrained single-channel foundation models, such as WavLM, achieved state-of-the-art performance. This work compares three approaches to integrate spatial features into foundation model-based diarization systems: the cascade of a beamformer and a single-channel foundation model, a multi-channel foundation model, and the conditioning of the downstream network on explicitly extracted spatial features. Results show that the beamformer front-end is even detrimental to diarization performance in regions of overlapped speech, while best performance is achieved with the conditioning, demonstrating that the incorporation of explicit spatial features is a competitive approach to foundation-model-supported diarization. This approach is further subjected to a detailed error analysis showing that the conditioning system removes errors to a good extent that would occur when either only spectral or only spatial features were used.

eess.AS↗

arXiv2026

Audio Diarization: A New Paradigm for Exploring Audio Recordings with Unknown Event Classes

We propose a new task, audio diarization. The motivation is that there are applications, such as audio monitoring in an unknown environment, where initially the sound event classes to be recognized are unknown. For such a scenario, we propose to first localize in time relevant sound events and to classify them, e.g., by comparing with known event classes, in a second step. This contribution is dedicated to the first step, which we call audio diarization, as it is reminiscent of the speaker diarization stage that precedes and simplifies the second stage, speech recognition, in multi-talker conversational speech processing. In this contribution, we define audio diarization as detecting onset and offset times of sound events with overlap for an open set of classes and without user prompts. We show how a speaker diarization system can be adjusted for audio diarization and propose an evaluation setup. Compared to a closed-set sound event detection system, the proposed system achieves similar performance with the additional ability to detect novel sounds.

eess.AS↗

arXiv2026

Speech Quality Embeddings for Improved Detection and Classification of Degradations in Speech Signals

Automatic subjective speech quality assessment (SSQA) traditionally estimates speech quality on an utterance or system level. While this resolution was adequate for older transmission or synthesis systems that produced speech signals of mediocre quality, modern systems generate high-quality speech with degradations that may occur only locally. With suitable model architectures and regularization losses, SSQA models trained with utterance-level targets can also yield useful local predictions of speech quality. In this work, we extend such models to produce frame-level embeddings that cluster by degradation type. Specifically, we employ a partial mix-up strategy on a parallel corpus of clean and degraded utterances and apply a contrastive loss to distinguish between degradation types. Through experiments on both in- and out-of-domain data, we demonstrate that our approach improves degradation detection and enables the identification of degradation types by analyzing embedding clusters.

eess.AS↗

arXiv2026

Disentangling Pitch and Creak for Speaker Identity Preservation in Speech Synthesis

We introduce a system capable of faithfully modifying the perceptual voice quality of creak while preserving the speaker's perceived identity. While it is well known that high creak probability is typically correlated with low pitch, it is important to note that this is a property observed on a population of speakers but does not necessarily hold across all situations. Disentanglement of pitch from creak is achieved by augmentation of the training dataset of a speech synthesis system with a speaker manipulation block based on conditional continuous normalizing flow. The experiments show greatly improved speaker verification performance over a range of creak manipulation strengths.

eess.AS↗

arXiv2026

Speech Quality-Based Localization of Low-Quality Speech and Text-to-Speech Synthesis Artefacts

A large number of works view the automatic assessment of speech from an utterance- or system-level perspective. While such approaches are good in judging overall quality, they cannot adequately explain why a certain score was assigned to an utterance. frame-level scores can provide better interpretability, but models predicting them are harder to tune and regularize since no strong targets are available during training. In this work, we show that utterance-level speech quality predictors can be regularized with a segment-based consistency constraint which notably reduces frame-level stochasticity. We then demonstrate two applications involving frame-level scores: The partial spoof scenario and the detection of synthesis artefacts in two state-of-the-art text-to-speech systems. For the latter, we perform listening tests and confirm that listeners rate segments to be of poor quality more often in the set defined by low frame-level scores than in a random control set.

eess.AS↗

arXiv2026

Loose coupling of spectral and spatial models for multi-channel diarization and enhancement of meetings in dynamic environments

Sound capture by microphone arrays opens the possibility to exploit spatial, in addition to spectral, information for diarization and signal enhancement, two important tasks in meeting transcription. However, there is no one-to-one mapping of positions in space to speakers if speakers move. Here, we address this by proposing a novel joint spatial and spectral mixture model, whose two submodels are loosely coupled by modeling the relationship between speaker and position index probabilistically. Thus, spatial and spectral information can be jointly exploited, while at the same time allowing for speakers speaking from different positions. Experiments on the LibriCSS data set with simulated speaker position changes show great improvements over tightly coupled subsystems.

eess.AS↗

arXiv2026

On the Role of Spatial Features in Foundation-Model-Based Speaker Diarization

Recent advances in speaker diarization exploit large pretrained foundation models, such as WavLM, to achieve state-of-the-art performance on multiple datasets. Systems like DiariZen leverage these rich single-channel representations, but are limited to single-channel audio, preventing the use of spatial cues available in multi-channel recordings. This work analyzes the impact of incorporating spatial information into a state-of-the-art single-channel diarization system by evaluating several strategies for conditioning the model on multi-channel spatial features. Experiments on meeting-style datasets indicate that spatial information can improve diarization performance, but the overall improvement is smaller than expected for the proposed system, suggesting that the features aggregated over all WavLM layers already capture much of the information needed for accurate speaker discrimination, also in overlapping speech regions. These findings provide insight into the potential and limitations of using spatial cues to enhance foundation model-based diarization.

eess.AS↗

arXiv2025

Synthesizing speech with selected perceptual voice qualities - A case study with creaky voice

The control of perceptual voice qualities in a text-to-speech (TTS) system is of interest for applications where unmanipu- lated and manipulated speech probes can serve to illustrate pho- netic concepts that are otherwise difficult to grasp. Here, we show that a TTS system, that is augmented with a global speaker attribute manipulation block based on normalizing flows1 , is capable of correctly manipulating the non-persistent, localized quality of creaky voice, thus avoiding the necessity of a, typi- cally unreliable, frame-wise creak predictor. Subjective listen- ing tests confirm successful creak manipulation at a slightly re- duced MOS score compared to the original recording.

eess.AS↗

arXiv2025

Towards Frame-level Quality Predictions of Synthetic Speech

While automatic subjective speech quality assessment has witnessed much progress, an open question is whether an automatic quality assessment at frame resolution is possible. This would be highly desirable, as it adds explainability to the assessment of speech synthesis systems. Here, we take first steps towards this goal by identifying issues of existing quality predictors that prevent sensible frame-level prediction. Further, we define criteria that a frame-level predictor should fulfill. We also suggest a chunk-based processing that avoids the impact of a localized distortion on the score of neighboring frames. Finally, we measure in experiments with localized artificial distortions the localization performance of a set of frame-level quality predictors and show that they can outperform detection performance of human annotations obtained from a crowd-sourced perception experiment.

eess.AS↗

arXiv2025

Error Analysis in a Modular Meeting Transcription System

Meeting transcription is a field of high relevance and remarkable progress in recent years. Still, challenges remain that limit its performance. In this work, we extend a previously proposed framework for analyzing leakage in speech separation with proper sensitivity to temporal locality. We show that there is significant leakage to the cross channel in areas where only the primary speaker is active. At the same time, the results demonstrate that this does not affect the final performance much as these leaked parts are largely ignored by the voice activity detection (VAD). Furthermore, different segmentations are compared showing that advanced diarization approaches are able to reduce the gap to oracle segmentation by a third compared to a simple energy-based VAD. We additionally reveal what factors contribute to the remaining difference. The results represent state-of-the-art performance on LibriCSS among systems that train the recognition module on LibriSpeech data only.

eess.AS↗

arXiv2025

Spatio-spectral diarization of meetings by combining TDOA-based segmentation and speaker embedding-based clustering

We propose a spatio-spectral, combined model-based and data-driven diarization pipeline consisting of TDOA-based segmentation followed by embedding-based clustering. The proposed system requires neither access to multi-channel training data nor prior knowledge about the number or placement of microphones. It works for both a compact microphone array and distributed microphones, with minor adjustments. Due to its superior handling of overlapping speech during segmentation, the proposed pipeline significantly outperforms the single-channel pyannote approach, both in a scenario with a compact microphone array and in a setup with distributed microphones. Additionally, we show that, unlike fully spatial diarization pipelines, the proposed system can correctly track speakers when they change positions.

eess.AS↗

arXiv2025

On the Application of Diffusion Models for Simultaneous Denoising and Dereverberation

Diffusion models have been shown to achieve natural-sounding enhancement of speech degraded by noise or reverberation. However, their simultaneous denoising and dereverberation capability has so far not been studied much, although this is arguably the most common scenario in a practical application. In this work, we investigate different approaches to enhance noisy and/or reverberant speech. We examine the cascaded application of models, each trained on only one of the distortions, and compare it with a single model, trained either solely on data that is both noisy and reverberated, or trained on data comprising subsets of purely noisy, of purely reverberated, and of noisy reverberant speech. Tests are performed both on artificially generated and real recordings of noisy and/or reverberant data. The results show that, when using the cascade of models, satisfactory results are only achieved if they are applied in the order of the dominating distortion. If only a single model is desired that can operate on all distortion scenarios, the best compromise appears to be a model trained on the aforementioned three subsets of degraded speech data.

eess.AS↗

arXiv2025

Word Error Rate Definitions and Algorithms for Long-Form Multi-talker Speech Recognition

The predominant metric for evaluating speech recognizers, the Word Error Rate (WER) has been extended in different ways to handle transcripts produced by long-form multi-talker speech recognizers. These systems process long transcripts containing multiple speakers and complex speaking patterns so that the classical WER cannot be applied. There are speaker-attributed approaches that count speaker confusion errors, such as the concatenated minimum-permutation WER cpWER and the time-constrained cpWER (tcpWER), and speaker-agnostic approaches, which aim to ignore speaker confusion errors, such as the Optimal Reference Combination WER (ORC-WER) and the MIMO-WER. These WERs evaluate different aspects and error types (e.g., temporal misalignment). A detailed comparison has not been made. We therefore present a unified description of the existing WERs and highlight when to use which metric. To further analyze how many errors are caused by speaker confusion, we propose the Diarization-invariant cpWER (DI-cpWER). It ignores speaker attribution errors and its difference to cpWER reflects the impact of speaker confusions on the WER. Since error types cannot reliably be classified automatically, we discuss ways to visualize sequence alignments between the reference and hypothesis transcripts to facilitate the spotting of errors by a human judge. Since some WER definitions have high computational complexity, we introduce a greedy algorithm to approximate the ORC-WER and DI-cpWER with high precision ($<0.1\%$ deviation in our experiments) and polynomial complexity instead of exponential. To improve the plausibility of the metrics, we also incorporate the time constraint from the tcpWER into ORC-WER and MIMO-WER, also significantly reducing the computational complexity.

eess.AS↗

arXiv2025

Combining TF-GridNet and Mixture Encoder for Continuous Speech Separation for Meeting Transcription

Many real-life applications of automatic speech recognition (ASR) require processing of overlapped speech. A common method involves first separating the speech into overlap-free streams on which ASR is performed. Recently, TF-GridNet has shown impressive performance in speech separation in real reverberant conditions. Furthermore, a mixture encoder was proposed that leverages the mixed speech to mitigate the effect of separation artifacts. In this work, we extended the mixture encoder from a static two-speaker scenario to a natural meeting context featuring an arbitrary number of speakers and varying degrees of overlap. We further demonstrate its limits by the integration with separators of varying strength including TF-GridNet. Our experiments result in a new state-of-the-art performance on LibriCSS using a single microphone. They show that TF-GridNet largely closes the gap between previous methods and oracle separation independent of mixture encoding. We further investigate the remaining potential for improvement.

eess.AS↗

arXiv2025

Simultaneous Diarization and Separation of Meetings through the Integration of Statistical Mixture Models

We propose an approach for simultaneous diarization and separation of meeting data. It consists of a complex Angular Central Gaussian Mixture Model (cACGMM) for speech source separation, and a von-Mises-Fisher Mixture Model (VMFMM) for diarization in a joint statistical framework. Through the integration, both spatial and spectral information are exploited for diarization and separation. We also develop a method for counting the number of active speakers in a segment of a meeting to support block-wise processing. While the total number of speakers in a meeting may be known, it is usually not known on a per-segment level. With the proposed speaker counting, joint diarization and source separation can be done segment-by-segment, and the permutation problem across segments is solved, thus allowing for block-online processing in the future. Experimental results on the LibriCSS meeting corpus show that the integrated approach outperforms a cascaded approach of diarization and speech enhancement in terms of WER, both on a per-segment and on a per-meeting level.

eess.AS↗

arXiv2025

30+ Years of Source Separation Research: Achievements and Future Challenges

Source separation (SS) of acoustic signals is a research field that emerged in the mid-1990s and has flourished ever since. On the occasion of ICASSP's 50th anniversary, we review the major contributions and advancements in the past three decades in the speech, audio, and music SS research field. We will cover both single- and multi-channel SS approaches. We will also look back on key efforts to foster a culture of scientific evaluation in the research field, including challenges, performance metrics, and datasets. We will conclude by discussing current trends and future research directions.

eess.AS↗

arXiv2025

Speech Synthesis along Perceptual Voice Quality Dimensions

While expressive speech synthesis or voice conversion systems mainly focus on controlling or manipulating abstract prosodic characteristics of speech, such as emotion or accent, we here address the control of perceptual voice qualities (PVQs) recognized by phonetic experts, which are speech properties at a lower level of abstraction. The ability to manipulate PVQs can be a valuable tool for teaching speech pathologists in training or voice actors. In this paper, we integrate a Conditional Continuous-Normalizing-Flow-based method into a Text-to-Speech system to modify perceptual voice attributes on a continuous scale. Unlike previous approaches, our system avoids direct manipulation of acoustic correlates and instead learns from examples. We demonstrate the system's capability by manipulating four voice qualities: Roughness, breathiness, resonance and weight. Phonetic experts evaluated these modifications, both for seen and unseen speaker conditions. The results highlight both the system's strengths and areas for improvement.

eess.AS↗

arXiv2025

Microphone Array Signal Processing and Deep Learning for Speech Enhancement

Multi-channel acoustic signal processing is a well-established and powerful tool to exploit the spatial diversity between a target signal and non-target or noise sources for signal enhancement. However, the textbook solutions for optimal data-dependent spatial filtering rest on the knowledge of second-order statistical moments of the signals, which have traditionally been difficult to acquire. In this contribution, we compare model-based, purely data-driven, and hybrid approaches to parameter estimation and filtering, where the latter tries to combine the benefits of model-based signal processing and data-driven deep learning to overcome their individual deficiencies. We illustrate the underlying design principles with examples from noise reduction, source separation, and dereverberation.

eess.AS↗