TY - CHAP A1 - Baumann, Ilja A1 - Wagner, Dominik A1 - Bayerl, Sebastian P. A1 - Bocklet, Tobias T1 - Nonwords Pronunciation Classification in Language Development Tests for Preschool Children T2 - Interspeech 2022 N2 - This work aims to automatically evaluate whether the language development of children is age-appropriate. Validated speech and language tests are used for this purpose to test the auditory memory. In this work, the task is to determine whether spoken nonwords have been uttered correctly. We compare different approaches that are motivated to model specific language structures: Low-level features (FFT), speaker embeddings (ECAPA-TDNN), grapheme-motivated embeddings (wav2vec 2.0), and phonetic embeddings in form of senones (ASR acoustic model). Each of the approaches provides input for VGG-like 5-layer CNN classifiers. We also examine the adaptation per nonword. The evaluation of the proposed systems was performed using recordings from different kindergartens of spoken nonwords. ECAPA-TDNN and low-level FFT features do not explicitly model phonetic information; wav2vec2.0 is trained on grapheme labels, our ASR acoustic model features contain (sub-)phonetic information. We found that the more granular the phonetic modeling is, the higher are the achieved recognition rates. The best system trained on ASR acoustic model features with VTLN achieved an accuracy of 89.4% and an area under the ROC (Receiver Operating Characteristic) curve (AUC) of 0.923. This corresponds to an improvement in accuracy of 20.2% and AUC of 0.309 relative compared to the FFT-baseline. Y1 - 2022 U6 - https://doi.org/10.21437/interspeech.2022-10777 SN - 2958-1796 VL - 2022 SP - 3643 EP - 3647 PB - ISCA ER - TY - JOUR A1 - Bayerl, Sebastian P. A1 - Wagner, Dominik A1 - Baumann, Ilja A1 - Bocklet, Tobias A1 - Riedhammer, Korbinian T1 - Detecting Vocal Fatigue with Neural Embeddings JF - Journal of Voice N2 - Vocal fatigue refers to the feeling of tiredness and weakness of voice due to extended utilization. This paper investigates the effectiveness of neural embeddings for the detection of vocal fatigue. We compare x-vectors, ECAPA-TDNN, and wav2vec 2.0 embeddings on a corpus of academic spoken English. Low-dimensional mappings of the data reveal that neural embeddings capture information about the change in vocal characteristics of a speaker during prolonged voice usage. We show that vocal fatigue can be reliably predicted using all three types of neural embeddings after 40 minutes of continuous speaking when temporal smoothing and normalization are applied to the extracted embeddings. We employ support vector machines for classification and achieve accuracy scores of 81% using x-vectors, 85% using ECAPA-TDNN embeddings, and 82% using wav2vec 2.0 embeddings as input features. We obtain an accuracy score of 76%, when the trained system is applied to a different speaker and recording environment without any adaptation. KW - Vocal fatigue KW - Neural embeddings KW - Visualization KW - Detection Y1 - 2023 U6 - https://doi.org/10.1016/j.jvoice.2023.01.012 SN - 0892-1997 PB - Elsevier BV ER - TY - CHAP A1 - Wagner, Dominik A1 - Bayerl, Sebastian P. A1 - Baumann, Ilja A1 - Riedhammer, Korbinian A1 - Nöth, Elmar A1 - Bocklet, Tobias T1 - Large Language Models for Dysfluency Detection in Stuttered Speech N2 - Accurately detecting dysfluencies in spoken language can help to improve the performance of automatic speech and language processing components and support the development of more inclusive speech and language technologies. Inspired by the recent trend towards the deployment of large language models (LLMs) as universal learners and processors of non-lexical inputs, such as audio and video, we approach the task of multi-label dysfluency detection as a language modeling problem. We present hypotheses candidates generated with an automatic speech recognition system and acoustic representations extracted from an audio encoder model to an LLM, and finetune the system to predict dysfluency labels on three datasets containing English and German stuttered speech. The experimental results show that our system effectively combines acoustic and lexical information and achieves competitive results on the multi-label stuttering detection task. KW - dysfluency detection, stuttering, large language models, wav2vec 2.0, Whisper, pathological speech Y1 - 2024 U6 - https://doi.org/10.48550/arXiv.2406.11025 ER - TY - CHAP A1 - Baumann, Ilja A1 - Unger, Nicole A1 - Wagner, Dominik A1 - Riedhammer, Korbinian A1 - Bocklet, Tobias T1 - Automatic Evaluation of a Sentence Memory Test for Preschool Children N2 - Assessment of memory capabilities in preschool-aged children is crucial for early detection of potential speech development impairments or delays. We present an approach for the automatic evaluation of a standardized sentence memory test specifically for preschool children. Our methodology leverages automatic transcription of recited sentences and evaluation based on natural language processing techniques. We demonstrate the effectiveness of our approach on a dataset comprised of recited sentences from preschool-aged children, incorporating ratings of semantic and syntactic correctness. The best performing systems achieve an F1 score of 91.7% for semantic correctness and 86.1% for syntactic correctness using automatic transcripts. Our results showcase the potential of automated evaluation systems in providing reliable and efficient assessments of memory capabilities in early childhood, facilitating timely interventions and support for children with language development needs. KW - speech development, children’s speech, automatic assessment Y1 - 2024 U6 - https://doi.org/10.21437/Interspeech.2024-2125 SP - 5158 EP - 5162 ER - TY - CHAP A1 - Baumann, Ilja A1 - Wagner, Dominik A1 - Schuster, Maria A1 - Riedhammer, Korbinian A1 - Nöth, Elmar A1 - Bocklet, Tobias T1 - Towards Self-Attention Understanding for Automatic Articulatory Processes Analysis in Cleft Lip and Palate Speech N2 - Cleft lip and palate (CLP) speech presents unique challenges for automatic phoneme analysis due to its distinct acoustic characteristics and articulatory anomalies. We perform phoneme analysis in CLP speech using a pre-trained wav2vec 2.0 model with a multi-head self-attention classification module to capture long-range dependencies within the speech signal, thereby enabling better contextual understanding of phoneme sequences. We demonstrate the effectiveness of our approach in the classification of various articulatory processes in CLP speech. Furthermore, we investigate the interpretability of self-attention to gain insights into the model’s understanding of CLP speech characteristics. Our findings highlight the potential of the selfattention mechanisms for improving automatic phoneme analysis in CLP speech, paving the way for enhanced diagnostics, adding interpretability for therapists and affected patients. KW - pathologic speech, cleft lip and palate, children’s speech, automatic assessment Y1 - 2024 U6 - https://doi.org/10.21437/Interspeech.2024-2134 SP - 2430 EP - 2434 ER - TY - CHAP A1 - Wagner, Dominik A1 - Baumann, Ilja A1 - Riedhammer, Korbinian A1 - Bocklet, Tobias ED - Pesak, Krisztina T1 - Outlier Reduction with Gated Attention for Improved Post-training Quantization in Large Sequence-to-sequence Speech Foundation Models N2 - This paper explores the improvement of post-training quantization (PTQ) after knowledge distillation in the Whisper speech foundation model family. We address the challenge of outliers in weights and activation tensors, known to impede quantization quality in transformer-based language and vision models. Extending this observation to Whisper, we demonstrate that these outliers are also present when transformer-based models are trained to perform automatic speech recognition, necessitating mitigation strategies for PTQ. We show that outliers can be reduced by a recently proposed gating mechanism in the attention blocks of the student model, enabling effective 8-bit quantization, and lower word error rates compared to student models without the gating mechanism in place. KW - post-training quantization, Whisper, gated attention, outliers Y1 - 2024 U6 - https://doi.org/10.21437/Interspeech.2024-2105 SP - 4623 EP - 4627 ER - TY - JOUR A1 - Escobar-Grisales, Daniel A1 - Ríos-Urrego, Cristian-David A1 - Baumann, Ilja A1 - Riedhammer, Korbinian A1 - Nöth, Elmar A1 - Bocklet, Tobias A1 - Garcia, Adolfo A1 - Orozco-Arroyave, Juan rafael T1 - It’s Time to Take Action: Acoustic Modeling of Motor Verbs to Detect Parkinson’s Disease N2 - Pre-trained models generate speech representations that are used in different tasks, including the automatic detection of Parkinson’s disease (PD). Although these models can yield high accuracy, their interpretation is still challenging. This paper used a pre-trained Wav2vec 2.0 model to represent speech frames of 25ms length and perform a frame-by-frame discrimination between PD patients and healthy control (HC) subjects. This fine granularity prediction enabled us to identify specific linguistic segments with high discrimination capability. Speech representations of all produced verbs were compared w.r.t. nouns and the first ones yielded higher accuracies. To gaina deeper understanding of this pattern, representations of motor and non-motor verbs were compared and the first ones yielded better results, with accuracies of around 83% in an independent test set. These findings support well-established neurocognitive models about action-related language highlighted as key drivers of PD. Index Terms: computational paralinguistics, interpretability of pre-trained models, action verbs, Parkinson’s disease KW - computational paralinguistics, interpretability of pre-trained models, action verbs, Parkinson’s disease Y1 - 2024 U6 - https://doi.org/10.21437/Interspeech.2024-2205 ER - TY - CHAP A1 - Bundscherer, Maximilian A1 - Schmitt, Thomas A1 - Baumann, Ilja A1 - Bocklet, Tobias T1 - Digital Operating Mode Classification of Real-World Amateur Radio Transmissions T2 - ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) N2 - This study presents an ML approach for classifying digital radio operating modes evaluated on real-world transmissions. We generated 98 different parameterized radio signals from 17 digital operating modes, transmitted each of them on the 70 cm (UHF) amateur radio band, and recorded our transmissions with two different architectures of SDR receivers. Three lightweight ML models were trained exclusively on spectrograms of limited non-transmitted signals with random characters as payloads. This training involved an online data augmentation pipeline to simulate various radio channel impairments. Our best model, EfficientNetB0, achieved an accuracy of 93.80% across the 17 operating modes and 85.47% across all 98 parameterized radio signals, evaluated on our real-world transmissions with Wikipedia articles as payloads. Furthermore, we analyzed the impact of varying signal durations & the number of FFT bins on classification, assessed the effectiveness of our simulated channel impairments, and tested our models across multiple simulated SNRs. KW - Automatic Modulation Classification KW - Amateur KW - Radio KW - Spectrum Monitoring KW - Cognitive Radio KW - Machine Learning Y1 - 2025 U6 - https://doi.org/10.1109/ICASSP49660.2025.10889837 SP - 1 EP - 5 PB - IEEE ER - TY - CHAP A1 - Baumann, Ilja A1 - Wagner, Dominik A1 - Riedhammer, Korbinian A1 - Bocklet, Tobias T1 - Optimized Self-supervised Training with BEST-RQ for Speech Recognition T2 - ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) N2 - Self-supervised learning has been successfully used for various speech related tasks, including automatic speech recognition. BERT-based Speech pre-Training with Random-projection Quantizer (BEST-RQ) has achieved state-of-the-art results in speech recognition. In this work, we further optimize the BEST-RQ approach using Kullback-Leibler divergence as an additional regularizing loss and multicodebook extension per cluster derived from low-level feature clustering. Preliminary experiments on train-100 split of LibriSpeech result in a relative improvement of 11.2% on test-clean by using multiple codebooks, utilizing a combination of cross-entropy and Kullback-Leibler divergence further reduces the word error rate by 4.5%. The proposed optimizations on full LibriSpeech pre-training and fine-tuning result in relative word error rate improvements of up to 23.8% on test-clean and 30.6% on testother using 6 codebooks. Furthermore, the proposed setup leads to faster convergence in pre-training and fine-tuning and additionally stabilizes the pre-training. KW - self-supervised learning KW - speech recognition KW - conformer KW - BEST-RQ Y1 - 2025 U6 - https://doi.org/10.1109/ICASSP49660.2025.10889362 SP - 1 EP - 5 PB - IEEE ER - TY - CHAP A1 - Wagner, Dominik A1 - Baumann, Ilja A1 - Engert, Natalie A1 - Lee, Seanie A1 - Nöth, Elmar A1 - Riedhammer, Korbinian A1 - Bocklet, Tobias T1 - Personalized Fine-Tuning with Controllable Synthetic Speech from LLM-Generated Transcripts for Dysarthric Speech Recognition T2 - Interspeech 2025 N2 - In this work, we present our submission to the Speech Accessibility Project challenge for dysarthric speech recognition. We integrate parameter-efficient fine-tuning with latent audio representations to improve an encoder-decoder ASR system. Synthetic training data is generated by fine-tuning Parler-TTS to mimic dysarthric speech, using LLM-generated prompts for corpus-consistent target transcripts. Personalization with x-vectors consistently reduces word error rates (WERs) over non-personalized fine-tuning. AdaLoRA adapters outperform full fine-tuning and standard low-rank adaptation, achieving relative WER reductions of ∼23% and ∼22%, respectively. Further improvements (∼5% WER reduction) come from incorporating wav2vec 2.0-based audio representations. Training with synthetic dysarthric speech yields up to ∼7% relative WER improvement over personalized fine-tuning alone. KW - speech recognition KW - dysarthric speech KW - personalization KW - low-rank adaptation KW - synthetic data Y1 - 2025 U6 - https://doi.org/10.21437/Interspeech.2025-2155 SN - 2958-1796 SP - 3294 EP - 3298 PB - ISCA CY - ISCA ER - TY - CHAP A1 - Baumann, Ilja A1 - Wagner, Dominik A1 - Riedhammer, Korbinian A1 - Bocklet, Tobias T1 - Pathology-Aware Speech Encoding and Data Augmentation for Dysarthric Speech Recognition T2 - Interspeech 2025 N2 - Automatic speech recognition (ASR) for pathologic speech remains a major challenge due to high variability in articulation, phonation, and prosody distortions. In this work, we propose a pathology-aware speech encoder based on BEST-RQ pre-training, which incorporates 46k hours of speech, including pathologic and atypical speech. We continue pre-training for domain adaptation and experiment with etiology-specific codebooks. We achieve a 13.2% relative word error rate (WER) improvement using the pathology-aware speech encoder with etiology-specific continued pre-training. Additionally, we examine the impact of incorporating synthetic and out-of-domain (OOD) data to further enhance ASR performance. Synthetic data reduces WER by up to 8.7%, while OOD data improves WER by 12.2%. Finally, we introduce a semantic similaritybased data augmentation technique to optimize data selection, achieving a WER improvement of up to 9.7% while minimizing the need for additional training data. KW - pathologic speech recognition KW - speech augmentation KW - self-supervised learning Y1 - 2025 U6 - https://doi.org/10.21437/Interspeech.2025-2724 SN - 2958-1796 SP - 3289 EP - 3293 PB - ISCA CY - ISCA ER - TY - CHAP A1 - Wagner, Dominik A1 - Baumann, Ilja A1 - Bocklet, Tobias T1 - Vocoder-Free Non-parallel Conversion of Whispered Speech With Masked Cycle-Consistent Generative Adversarial Networks N2 - Cycle-consistent generative adversarial networks have been widely used in non-parallel voice conversion (VC). Their ability to learn mappings between source and target features without relying on parallel training data eliminates the need for temporal alignments. However, most methods decouple the conversion of acoustic features from synthesizing the audio signal by using separate models for conversion and waveform synthesis. This work unifies conversion and synthesis into a single model, thereby eliminating the need for a separate vocoder. By leveraging cycle-consistent training and a self-supervised auxiliary training task, our model is able to efficiently generate converted high-quality raw audio waveforms. Subjective listening tests showed that our unified approach achieved improvements of up to 6.7% relative to the baseline in whispered VC. Mean opinion score predictions also yielded stable results in conventional VC (between 0.5% and 2.4% relative improvement). KW - voice conversion KW - generative adversarial networks KW - cycle-consistency KW - masking KW - whispered speech Y1 - 2025 SN - 978-3-032-02548-7 U6 - https://doi.org/10.1007/978-3-032-02548-7_20 PB - Springer CY - Cham ER -