Multimedia event extraction aims to jointly identify events and their arguments across multiple modalities, such as text and images, to support more comprehensive event understanding.While recent work reports steady and substantial progress, the reliability and comparability of these results critically depend on consistent and rigorous evaluation.In this work, we present the first systematic analysis of evaluation pitfalls in multimedia event extraction and identify three major sources of issues: inconsistent data processing, inconsistent task assumptions, and overly relaxed evaluation settings.We demonstrate, through a series of controlled experiments under a strict evaluation framework, that minor evaluation choices can cause large performance variations and lead to overestimation of a model's ability to ground real-world events across modalities.Our findings highlight the need for comparable evaluation standards and encourage a shift toward more rigorous evaluation in multimedia event extraction.

Early and accessible detection of Alzheimer's disease (AD) remains a major challenge, as current diagnostic methods often rely on costly and invasive biomarkers. Speech and language analysis has emerged as a promising non-invasive and scalable approach to detecting cognitive impairment, but research in this area is hindered by the lack of publicly available datasets, especially for languages other than English. This paper introduces the PARLO Dementia Corpus (PDC), a new multi-center, clinically validated German resource for AD collected across nine academic memory clinics in Germany. The dataset comprises speech recordings from individuals with AD-related mild cognitive impairment and mild to moderate dementia, as well as cognitively healthy controls. Speech was elicited using a standardized test battery of eight neuropsychological tasks, including confrontation naming, verbal fluency, word repetition, picture description, story reading, and recall tasks. In addition to audio recordings, the dataset includes manually verified transcriptions and detailed demographic, clinical, and biomarker metadata. Baseline experiments on ASR benchmarking, automated test evaluation, and LLM-based classification illustrate the feasibility of automatic, speech-based cognitive assessment and highlight the diagnostic value of recall-driven speech production. The PDC thus establishes the first publicly available German benchmark for multi-modal and cross-lingual research on neurodegenerative diseases.

Assessing spoken language abilities in children is critical for early detection of developmental delays. We propose a two-stage framework for automatic pronunciation assessment, targeting classification of items in a language development screening. In the first stage, we fine-tune a sentence transformer using phonetic and label-informed losses to create a structured latent text space. In the second stage, we fine-tune an audio encoder to project spoken utterances into the text space. Our approach enables inference from audio alone while preserving alignment with phonetically meaningful structure. Compared to baseline systems, our method achieves the best classification performance and provides a latent space suitable for downstream analysis. This facilitates the identification of phonological or grammatical error patterns from cluster structure, paving the way for potentially interpretable assessment of early language acquisition.

Automatic transcription of stuttered speech remains a challenge, even for modern end-to-end (E2E) automatic speech recognition (ASR) frameworks. Dysfluencies and fluency-shaping artifacts are often overlooked, resulting in non-verbatim transcriptions with limited clinical and research value. We propose a parameter-efficient adaptation method to decode dysfluencies and fluency modifications as special tokens within transcriptions, evaluated on simulated (LibriStutter, English) and natural (KSoF, German) stuttered speech datasets. To mitigate ASR performance disparities and bias towards English, we introduce a multi-step fine-tuning strategy with language-adaptive pretraining. Tokenization analysis further highlights the tokenizer’s English-centric bias, which poses challenges for improving performance on German data. Our findings demonstrate the effectiveness of lightweight adaptation techniques for dysfluency-aware ASR while exposing key limitations in multilingual E2E systems.

We enhance the MaViLS multimodal algorithm to improve speech-to-slide alignment for lecture podcasts. Our approach integrates vision large language models for optical character recognition and automatic generation of lecture transcripts from slide content, coupled with a multilingual multimodal embedding model for text and image alignment. By combining slide-extracted text with automatically generated lecture transcripts and captioned slide images, we generate enhanced audio features that better capture speech-slide mapping. Our method improves the average F1 score for audio feature alignment on the English MaViLS dataset from 0.51 to 0.71 and on a newly created German podcast lectures dataset from 0.65 to 0.84.

Dysarthric speech recognition systems often focus solely on transcription, limiting their applicability in clinical and assistive settings where assessments of perceptual attributes like intelligibility and naturalness are essential. We propose a multimodal conversational framework based on Phi-4-Multimodal that combines ASR with attribute rating prediction, enabling users to query both transcriptions and perceptual characteristics (e.g. “How intelligible is this utterance?”). Our multitask model adds auxiliary prediction heads for five clinically relevant attributes and is trained on the English Speech Accessibility Project dataset. The system achieves competitive ASR performance while delivering attribute-level feedback comparable to specialized classifiers. Additional experiments show improved ASR performance for German Parkinson’s speech, indicating preserved multilingual capabilities and partial cross-lingual transfer of dysarthric speech patterns.

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 multi-codebook 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 test-other using 6 codebooks. Furthermore, the proposed setup leads to faster convergence in pre-training and fine-tuning and additionally stabilizes the pre-training.

We present an approach to Audio-Visual Speech Recognition that builds on a pre-trained Whisper model. To infuse visual information into this audio-only model, we extend it with an AV fusion module and LoRa adapters, one of the most up-to-date adapter approaches. One advantage of adapter-based approaches, is that only a relatively small number of parameters are trained, while the basic model remains unchanged. Common AVSR approaches train single models to handle several noise categories and noise levels simultaneously. Taking advantage of the lightweight nature of adapter approaches, we train noise-scenario-specific adapter-sets, each covering individual noise-categories or a specific noise-level range. The most suitable adapter-set is selected by previously classifying the noise-scenario. This enables our models to achieve an optimum coverage across different noise-categories and noise-levels, while training only a minimum number of parameters.Compared to a full fine-tuning approach with SOTA performance our models achieve almost comparable results over the majority of the tested noise-categories and noise-levels, with up to 88.5% less trainable parameters. Our approach can be extended by further noise-specific adapter-sets to cover additional noise scenarios. It is also possible to utilize the underlying powerful ASR model when no visual information is available, as it remains unchanged.

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.

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.

Current work on speech-based dementia assessment focuses on either feature extraction to predict assessment scales, or on the automation of existing test procedures. Most research uses public data unquestioningly and rarely performs a detailed error analysis, focusing primarily on numerical performance. We perform an in-depth analysis of an automated standardized dementia assessment, the Syndrom-Kurz-Test. We find that while there is a high overall correlation with human annotators, due to certain artifacts, we observe high correlations for the severely impaired individuals, which is less true for the healthy or mildly impaired ones. Speech production decreases with cognitive decline, leading to overoptimistic correlations when test scoring relies on word naming. Depending on the test design, fallback handling introduces further biases that favor certain groups. These pitfalls remain independent of group distributions in datasets and require differentiated analysis of target groups.

Segmenting speech transcripts into thematic sections benefits both downstream processing and users who depend on written text for accessibility. We introduce a novel approach to hierarchical topic segmentation in transcripts, generating multi-level tables of contents that capture both topic and subtopic boundaries. We compare zero-shot prompting and LoRA fine-tuning on large language models, while also exploring the integration of high-level speech pause features. Evaluations on English meeting recordings and multilingual lecture transcripts (Portuguese, German) show significant improvements over established topic segmentation baselines. Additionally, we adapt a common evaluation measure for multi-level segmentation, taking into account all hierarchical levels within one metric.

In this work, we optimize speculative sampling for parallel hardware accelerators to improve sampling speed.We notice that substantial portions of the intermediate matrices necessary for speculative sampling can be computed concurrently.This allows us to distribute the workload across multiple GPU threads, enabling simultaneous operations on matrix segments within thread blocks.This results in profiling time improvements ranging from 6% to 13% relative to the baseline implementation, without compromising accuracy.To further accelerate speculative sampling, probability distributions parameterized by softmax are approximated by sigmoid.This approximation approach results in significantly greater relative improvements in profiling time, ranging from 37% to 94%, with a minor decline in accuracy.We conduct extensive experiments on both automatic speech recognition and summarization tasks to validate the effectiveness of our optimization methods.

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.

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.

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.

Speech pauses, alongside content and structure, offer a valuable and non-invasive biomarker for detecting dementia. This work investigates the use of pause-enriched transcripts in transformer-based language models to differentiate the cognitive states of subjects with no cognitive impairment, mild cognitive impairment, and Alzheimer's dementia based on their speech from a clinical assessment. We address three binary classification tasks: Onset, monitoring, and dementia exclusion. The performance is evaluated through experiments on a German Verbal Fluency Test and a Picture Description Test, comparing the model's effectiveness across different speech production contexts. Starting from a textual baseline, we investigate the effect of incorporation of pause information and acoustic context. We show the test should be chosen depending on the task, and similarly, lexical pause information and acoustic cross-attention contribute differently.

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

We present a method to personalize large transformer-based encoder-decoder speech foundation models without the need for changes in the underlying model structure or training from scratch. This is achieved by projecting speaker-specific information into the latent space of the transformer decoder via a small neural network and learning to process the speaker information along with domain-specific information via parameter-efficient finetuning. We use this method to improve the automatic speech recognition results of spoken academic German and English. Our approach yields average relative word error rate (WER) improvements of approximately 29% on German academic speech and 25% on English academic speech. It also translates well to conversational speech, achieving relative WER improvements of up to 36%, and demonstrates modest gains of up to 5% on read speech. Moreover, we observe that incorporating utterances from the recent past as personalization context yields the most significant overall improvements and that changes in voice characteristics resulting from prolonged speaking have a minimal effect on the personalization quality of academic lectures.

The analysis of phonological processes is crucial in evaluating speech development disorders in children, but encounters challenges due to limited children audio data. This work focuses on automatic vowel error detection using a two-stage pipeline. The first stage uses a fine-tuned cross-lingual phone recognizer (wav2vec 2.0) to extract phone sequences from audio. The second stage employs a language model (BERT) for classification from a phone sequence, entirely trained on synthetic transcripts, to counteract the very broad range of potential mistakes. We evaluate the system on nonword audio recordings recited by preschool children from a speech development test. The results show that the classifier trained on synthetic data performs well, but its efficacy relies on the quality of the phone recognizer. The best classifier achieves an 94.7% F1 score when evaluated against phonetic ground truths, whereas the F1 score is 76.2% when using automatically recognized phone sequences.

We analyze the impact of speaker adaptation in end-to-end automatic speech recognition models based on transformers and wav2vec 2.0 under different noise conditions. By including speaker embeddings obtained from x-vector and ECAPA-TDNN systems, as well as i-vectors, we achieve relative word error rate improvements of up to 16.3% on LibriSpeech and up to 14.5% on Switchboard. We show that the proven method of concatenating speaker vectors to the acoustic features and supplying them as auxiliary model inputs remains a viable option to increase the robustness of end-to-end architectures. The effect on transformer models is stronger, when more noise is added to the input speech. The most substantial benefits for systems based on wav2vec 2.0 are achieved under moderate or no noise conditions. Both x-vectors and ECAPA-TDNN embeddings outperform i-vectors as speaker representations. The optimal embedding size depends on the dataset and also varies with the noise condition.

Progressive loss of memory is the most known symptom of Alzheimer’s Disease (AD); however, it also affects other cognitive skills and leads to depression symptoms. This paper presents a transfer learning strategy for automatically detecting AD and depression in AD patients using acoustic information and ForestNet, an artificial neural network that allows computing the contribution of a set of features to a model’s decision. The methodology consists of training ForestNet with a dataset commonly used for emotion recognition; then, we fine-tune the pre-trained model to detect AD and depression in AD. We trained the models with several acoustic features commonly used for emotion and AD applications. Unweighted average recalls of up to 0.87 were achieved to classify the disease and up to 0.82 to detect depression in AD. Our results indicate that the information obtained from the Arousal Valence plane may be suitable for discriminating and analyzing depression in AD.

Automated dementia screening enables early detection and intervention, reducing costs to healthcare systems and increasing quality of life for those affected. Depression has shared symptoms with dementia, adding complexity to diagnoses. The research focus so far has been on binary classification of dementia (DEM) and healthy controls (HC) using speech from picture description tests from a single dataset. In this work, we apply established baseline systems to discriminate cognitive impairment in speech from the semantic Verbal Fluency Test and the Boston Naming Test using text, audio and emotion embeddings in a 3-class classification problem (HC vs. MCI vs. DEM). We perform cross-corpus and mixed-corpus experiments on two independently recorded German datasets to investigate generalization to larger populations and different recording conditions. In a detailed error analysis, we look at depression as a secondary diagnosis to understand what our classifiers actually learn.

The detection of pathologies from speech features is usually defined as a binary classification task with one class representing a specific pathology and the other class representing healthy speech. In this work, we train neural networks, large margin classifiers, and tree boosting machines to distinguish between four pathologies: Parkinson's disease, laryngeal cancer, cleft lip and palate, and oral squamous cell carcinoma. We show that latent representations extracted at different layers of a pre-trained wav2vec 2.0 system can be effectively used to classify these types of pathological voices. We evaluate the robustness of our classifiers by adding room impulse responses to the test data and by applying them to unseen speech corpora. Our approach achieves unweighted average F1-Scores between 74.1% and 97.0%, depending on the model and the noise conditions used. The systems generalize and perform well on unseen data of healthy speakers sampled from a variety of different sources.

We are interested in the problem of conversational analysis and its application to the health domain. Cognitive Behavioral Therapy is a structured approach in psychotherapy, allowing the therapist to help the patient to identify and modify the malicious thoughts, behavior, or actions. This cooperative effort can be evaluated using the Working Alliance Inventory Observer-rated Shortened – a 12 items inventory covering task, goal, and relationship – which has a relevant influence on therapeutic outcomes. In this work, we investigate the relation between this alliance inventory and the spoken conversations (sessions) between the patient and the psychotherapist. We have delivered eight weeks of e-therapy, collected their audio and video call sessions, and manually transcribed them. The spoken conversations have been annotated and evaluated with WAI ratings by professional therapists. We have investigated speech and language features and their association with WAI items. The feature types include turn dynamics, lexical entrainment, and conversational descriptors extracted from the speech and language signals. Our findings provide strong evidence that a subset of these features are strong indicators of working alliance. To the best of our knowledge, this is the first and a novel study to exploit speech and language for characterising working alliance

Stuttering is a varied speech disorder that harms an individual's communication ability. Persons who stutter (PWS) often use speech therapy to cope with their condition. Improving speech recognition systems for people with such non-typical speech or tracking the effectiveness of speech therapy would require systems that can detect dysfluencies while at the same time being able to detect speech techniques acquired in therapy. This paper shows that fine-tuning wav2vec 2.0 [1] for the classification of stuttering on a sizeable English corpus containing stuttered speech, in conjunction with multi-task learning, boosts the effectiveness of the general-purpose wav2vec 2.0 features for detecting stuttering in speech; both within and across languages. We evaluate our method on FluencyBank , [2] and the German therapy-centric Kassel State of Fluency (KSoF) [3] dataset by training Support Vector Machine classifiers using features extracted from the finetuned models for six different stuttering-related event types: blocks, prolongations, sound repetitions, word repetitions, interjections, and - specific to therapy - speech modifications. Using embeddings from the fine-tuned models leads to relative classification performance gains up to 27% w.r.t. F1-score.

For dementia screening and monitoring, standardized tests play a key role in clinical routine since they aim at minimizing subjectivity by measuring performance on a variety of cognitive tasks. In this paper, we report a study consisting of a semistandardized history taking followed by two standardized neuropsychological tests, namely the SKT and the CERAD-NB. The tests include basic tasks such as naming objects, learning word lists, but also widely used tools such as the MMSE. Most of the tasks are performed verbally and should thus be suitable for automated scoring based on transcripts. For the first batch of 30 patients, we analyze the correlation between expert manual evaluations and automatic evaluations based on manual and automatic transcriptions. For both SKT and CERAD-NB, we observe high to perfect correlations using manual transcripts; for certain tasks with lower correlation, the automatic scoring is stricter than the human reference since it is limited to the audio. Using automatic transcriptions, correlations drop as expected and are related to recognition accuracy; however, we still observe high correlations of up to 0.98 (SKT) and 0.85 (CERADNB). We show that using word alternatives helps to mitigate recognition errors and subsequently improves correlation with expert scores.

The “Switchboard benchmark” is a very well-known test set in automatic speech recognition (ASR) research, establishing record-setting performance for systems that claim human-level transcription accuracy. This work highlights lesser-known practical considerations of this evaluation, demonstrating major improvements in word error rate (WER) by correcting the reference transcriptions and deviating from the official scoring methodology. In this more detailed and reproducible scheme, even commercial ASR systems can score below 5% WER and the established record for a research system is lowered to 2.3%. An alternative metric of transcript precision is proposed, which does not penalize deletions and appears to be more discriminating for human vs. machine performance. While commercial ASR systems are still below this threshold, a research system is shown to clearly surpass the accuracy of commercial human speech recognition. This work also explores using standardized scoring tools to compute oracle WER by selecting the best among a list of alternatives. A phrase alternatives representation is compared to utterance-level N-best lists and word-level data structures; using dense lattices and adding out-of-vocabulary words, this achieves an oracle WER of 0.18%.

Stuttering is a complex speech disorder that negatively affects an individual’s ability to communicate effectively. Persons who stutter (PWS) often suffer considerably under the condition and seek help through therapy. Fluency shaping is a therapy approach where PWSs learn to modify their speech to help them to overcome their stutter. Mastering such speech techniques takes time and practice, even after therapy. Shortly after therapy, success is evaluated highly, but relapse rates are high. To be able to monitor speech behavior over a long time, the ability to detect stuttering events and modifications in speech could help PWSs and speech pathologists to track the level of fluency. Monitoring could create the ability to intervene early by detecting lapses in fluency. To the best of our knowledge, no public dataset is available that contains speech from people who underwent stuttering therapy that changed the style of speaking. This work introduces the Kassel State of Fluency (KSoF), a therapy-based dataset containing over 5500 clips of PWSs. The clips were labeled with six stuttering-related event types: blocks, prolongations, sound repetitions, word repetitions, interjections, and – specific to therapy – speech modifications. The audio was recorded during therapy sessions at the Institut der Kasseler Stottertherapie. The data will be made available for research purposes upon request.

Personal Narrative (PN) is the recollection of individuals’ life experiences, events, and thoughts along with the associated emotions in the form of a story. Compared to other genres such as social media texts or microblogs, where people write about experienced events or products, the spoken PNs are complex to analyze and understand. They are usually long and unstructured, involving multiple and related events, characters as well as thoughts and emotions associated with events, objects, and persons. In spoken PNs, emotions are conveyed by changing the speech signal characteristics as well as the lexical content of the narrative. In this work, we annotate a corpus of spoken personal narratives, with the emotion valence using discrete values. The PNs are segmented into speech segments, and the annotators annotate them in the discourse context, with values on a 5-point bipolar scale ranging from -2 to +2 (0 for neutral). In this way, we capture the unfolding of the PNs events and changes in the emotional state of the narrator. We perform an in-depth analysis of the inter-annotator agreement, the relation between the label distribution w.r.t. the stimulus (positive/negative) used for the elicitation of the narrative, and compare the segment-level annotations to a baseline continuous annotation. We find that the neutral score plays an important role in the agreement. We observe that it is easy to differentiate the positive from the negative valence while the confusion with the neutral label is high. Keywords: Personal Narratives, Emotion Annotation, Segment Level Annotation

The ACM Multimedia 2022 Computational Paralinguistics Challenge addresses four different problems for the first time in a research competition under well-defined conditions: In the Vocalisations and Stuttering Sub-Challenges, a classification on human non-verbal vocalisations and speech has to be made; the Activity Sub-Challenge aims at beyond-audio human activity recognition from smartwatch sensor data; and in the Mosquitoes Sub-Challenge, mosquitoes need to be detected. We describe the Sub-Challenges, baseline feature extraction, and classifiers based on the 'usual' ComParE and BoAW features, the auDeep toolkit, and deep feature extraction from pre-trained CNNs using the DeepSpectrum toolkit; in addition, we add end-to-end sequential modelling, and a log-mel-128-BNN.

Personal narratives (PN) – spoken or written – are recollections of facts, people, events, and thoughts from one's own experience. Emotion recognition and sentiment analysis tasks are usually defined at the utterance or document level. However, in this work, we focus on Emotion Carriers (EC) defined as the segments (speech or text) that best explain the emotional state of the narrator (“loss of father”, “made me choose”). Once extracted, such EC can provide a richer representation of the user state to improve natural language understanding and dialogue modeling. In previous work, it has been shown that EC can be identified using lexical features. However, spoken narratives should provide a richer description of the context and the users' emotional state. In this paper, we leverage word-based acoustic and textual embeddings as well as early and late fusion techniques for the detection of ECs in spoken narratives. For the acoustic word-level representations, we use Residual Neural Networks (ResNet) pretrained on separate speech emotion corpora and fine-tuned to detect EC. Experiments with different fusion and system combination strategies show that late fusion leads to significant improvements for this task.

Alzheimer’s Disease (AD) results from the progressive loss of neurons in the hippocampus, which affects the capability to produce coherent language. It affects lexical, grammatical, and semantic processes as well as speech fluency. This paper considers the analyses of speech and language for the assessment of AD in the context of the Alzheimer’s Dementia Recognition through Spontaneous Speech (ADReSSo) 2021 challenge. We propose to extract acoustic features such as X-vectors, prosody, and emotional embeddings as well as linguistic features such as perplexity, and word-embeddings. The data consist of speech recordings from AD patients and healthy controls. The transcriptions are obtained using a commercial automatic speech recognition system. We outperform baseline results on the test set, both for the classification and the Mini-Mental State Examination (MMSE) prediction. We achieved a classification accuracy of 80% and an RMSE of 4.56 in the regression. Additionally, we found strong evidence for the influence of the interviewer on classification results. In cross-validation on the training set, we get classification results of 85% accuracy using the combined speech of the interviewer and the participant. Using interviewer speech only we still get an accuracy of 78%. Thus, we provide strong evidence for interviewer influence on classification results.

Performing machine learning tasks in mobile applications yields a challenging conflict of interest: highly sensitive client information (e.g., speech data) should remain private while also the intellectual property of service providers (e.g., model parameters) must be protected. Cryptographic techniques offer secure solutions for this, but have an unacceptable overhead and moreover require frequent network interaction.In this work, we design a practically efficient hardware-based solution. Specifically, we build OFFLINE MODEL GUARD (OMG) to enable privacy-preserving machine learning on the predominant mobile computing platform ARM—even in offline scenarios. By leveraging a trusted execution environment for strict hardware-enforced isolation from other system components, OMG guarantees privacy of client data, secrecy of provided models, and integrity of processing algorithms. Our prototype implementation on an ARM HiKey 960 development board performs privacy-preserving keyword recognition using TensorFlow Lite for Microcontrollers in real time.

With the advent of smart-home devices providing voice-based interfaces, such as Amazon Alexa or Apple Siri, voice data is constantly transferred to cloud services for automated speech recognition or speaker verification.While this development enables intriguing new applications, it also poses significant risks: Voice data is highly sensitive since it contains biometric information of the speaker as well as the spoken words.This data may be abused if not protected properly, thus the security and privacy of billions of end-users is at stake.We tackle this challenge by proposing an architecture, dubbed VoiceGuard, that efficiently protects the speech processing task inside a trusted execution environment (TEE).Our solution preserves the privacy of users while at the same time it does not require the service provider to reveal model parameters.Our architecture can be extended to enable user-specific models, such as feature transformations (including fMLLR), i-vectors, or model transformations (e.g., custom output layers).It also generalizes to secure on-premise solutions, allowing vendors to securely ship their models to customers.We provide a proof-of-concept implementation and evaluate it on the Resource Management and WSJ speech recognition tasks isolated with Intel SGX, a widely available TEE implementation, demonstrating even real time processing capabilities.

In this paper we present an algorithm that produces pitch and probability-of-voicing estimates for use as features in automatic speech recognition systems. These features give large performance improvements on tonal languages for ASR systems, and even substantial improvements for non-tonal languages. Our method, which we are calling the Kaldi pitch tracker (because we are adding it to the Kaldi ASR toolkit), is a highly modified version of the getf0 (RAPT) algorithm. Unlike the original getf0 we do not make a hard decision whether any given frame is voiced or unvoiced; instead, we assign a pitch even to unvoiced frames while constraining the pitch trajectory to be continuous. Our algorithm also produces a quantity that can be used as a probability of voicing measure; it is based on the normalized autocorrelation measure that our pitch extractor uses. We present results on data from various languages in the BABEL project, and show a large improvement over systems without tonal features and systems where pitch and POV information was obtained from SAcC or getf0.

In this paper we apply diagnostic analysis to gain a deeper understanding of the performance of the the keyword search system that we have developed for conversational telephone speech in the IARPA Babel program. We summarize the Babel task, its primary performance metric, “actual term weighted value” (ATWV), and our recognition and keyword search systems. Our analysis uses two new oracle ATWV measures, a bootstrap-based ATWV confidence interval, and includes a study of the underpinnings of the large ATWV gains due to system combination. This analysis quantifies the potential ATWV gains from improving the number of true hits and the overall quality of the detection scores in our system's posting lists. It also shows that system combination improves our systems' ATWV via a small increase in the number of true hits in the posting lists.

Cleft Lip and Palate (CLP) is among the most frequent congenital abnormalities. The impaired facial development affects the articulation, with different phonemes being impacted inhomogeneously among different patients. This work focuses on automatic phoneme analysis of children with CLP for a detailed diagnosis and therapy control. In clinical routine, the state-of-the-art evaluation is based on perceptual evaluations. Perceptual ratings act as ground-truth throughout this work, with the goal to build an automatic system that is as reliable as humans. We propose two different automatic systems focusing on modeling the articulatory space of a speaker: one system models a speaker by a GMM, the other system employs a speech recognition system and estimates fMLLR matrices for each speaker. SVR is then used to predict the perceptual ratings. We show that the fMLLR-based system is able to achieve automatic phoneme evaluation results that are in the same range as perceptual inter-rater-agreements.

We describe a lattice generation method that is exact, i.e. it satisfies all the natural properties we would want from a lattice of alternative transcriptions of an utterance. This method does not introduce substantial overhead above one-best decoding. Our method is most directly applicable when using WFST decoders where the WFST is “fully expanded”, i.e. where the arcs correspond to HMM transitions. It outputs lattices that include HMM-state-level alignments as well as word labels. The general idea is to create a state-level lattice during decoding, and to do a special form of determinization that retains only the best-scoring path for each word sequence. This special determinization algorithm is a solution to the following problem: Given a WFST A, compute a WFST B that, for each input-symbol-sequence of A, contains just the lowest-cost path through A.

In the past decade, semi-continuous hidden Markov models (SCHMMs) have not attracted much attention in the speech recognition community. Growing amounts of training data and increasing sophistication of model estimation led to the impression that continuous HMMs are the best choice of acoustic model. However, recent work on recognition of under-resourced languages faces the same old problem of estimating a large number of parameters from limited amounts of transcribed speech. This has led to a renewed interest in methods of reducing the number of parameters while maintaining or extending the modeling capabilities of continuous models. In this work, we compare classic and multiple-codebook semi-continuous models using diagonal and full covariance matrices with continuous HMMs and subspace Gaussian mixture models. Experiments on the RM and WSJ corpora show that while a classical semicontinuous system does not perform as well as a continuous one, multiple-codebook semi-continuous systems can perform better, particular when using full-covariance Gaussians.

Voice scrambling is widely used to add privacy to the radio communication of various authorities - but is also used by criminals to evade prosecution. In this article, we consider various analog voice scrambling techniques such as fixed frequency inversion, splitband inversion and rolling code scramblers. We explain how to break them using automatically extracted measures and scoring algorithms, and evaluate the proposed system using simulated data. While the simple inversion can be easily broken, the more advanced techniques require additional work prior to unsupervised automatization; the presented user interface allows the user to refine the automatic results to obtain a high quality solution.

In earlier studies, we employed a large prosodic feature vector to assess the quality of L2 learner's utterances with respect to sentence melody and rhythm.In this paper, we combine these features with two standard approaches in paralinguistic analysis: (1) features derived from a Gaussian Mixture Model used as Universal Background Model (GMM-UBM), and (2) openSMILE, an open-source toolkit for extracting acoustic features.We evaluate our approach with English speech from 94 non-native speakers perceptually scored by 62 native labellers.GMM-UBM or openSMILE modelling alone yields lower performance than our prosodic feature vector; however, adding information from the GMM-UBM modelling or openSMILE by late fusion improves results.

A growing number of universities and other educational institutions provide recordings of lectures and seminars as an additional resource to the students. In contrast to educational films that are scripted, directed and often shot by film professionals, these plain recordings are typically not post-processed in an editorial sense. Thus, the videos often contain longer periods of inactivity or silence, unnecessary repetitions, or corrections of prior mistakes. This paper describes the FAU Video Lecture Browser system, a web-based platform for the interactive assessment of video lectures, that helps to close the gap between a plain recording and a useful e-learning resource by displaying automatically extracted and ranked key phrases on an augmented time line based on stream graphs. In a pilot study, users of the interface were able to complete a topic localization task about 29 % faster than users provided with the video only while achieving about the same accuracy. The user interactions can be logged on the server to collect data to evaluate the quality of the phrases and rankings, and to train systems that produce customized phrase rankings.

In this work we focus on speaker verification on channels of varying quality, namely Skype and high frequency (HF) radio. In our setup, we assume to have telephone recordings of speakers for training, but recordings of different channels for testing with varying (lower) signal quality. Starting from a Gaussian mixture / support vector machine (GMM/SVM) baseline, we evaluate multi-condition training (MCT), an ideal channel classification approach (ICC), and nuisance attribute projection (NAP) to compensate for the loss of information due to the transmission. In an evaluation on Switchboard-2 data using Skype and HF channel simulators, we show that, for good signal quality, NAP improves the baseline system performance from 5% EER to 3.33% EER (for both Skype and HF). For strongly distorted data, MCT or, if adequate, ICC turn out to be the method of choice.

This paper focuses on the automatic detection of a person’s blood level alcohol based on automatic speech processing ap-proaches. We compare 5 different feature types with different ways of modeling. Experiments are based on the ALC corpus of IS2011 Speaker State Challenge. The classification task is restricted to the detection of a blood alcohol level above 0.5‰. Three feature sets are based on spectral observations: MFCCs, PLPs, TRAPS. These are modeled by GMMs. Classification is either done by a Gaussian classifier or by SVMs. In the later case classification is based on GMM-based supervectors, i.e. concatenation of GMM mean vectors. A prosodic system extracts a 292-dimensional feature vector based on a voiced-unvoiced decision. A transcription-based system makes use of text transcriptions related to phoneme durations and textual structure. We compare the stand-alone performances of these systems and combine them on score level by logistic regres-sion. The best stand-alone performance is the transcription-based system which outperforms the baseline by 4.8 % on the development set. A Combination on score level gave a huge boost when the spectral-based systems were added (73.6 %). This is a relative improvement of 12.7 % to the baseline. On the test-set we achieved an UA of 68.6 % which is a significant improvement of 4.1 % to the baseline system. Index Terms: GMM, alcohol intoxication, system fusion 1.

In this paper, we describe a new Java framework for an easy and efficient way of developing new GUI based speech processing applications. Standard components are provided to display the speech signal, the power plot, and the spectrogram. Furthermore, a component to create a new transcription and to display and manipulate an existing transcription is provided, as well as a component to display and manually correct external pitch values. These Swing components can be easily embedded into own Java programs. They can be synchronized to display the same region of the speech file. The object-oriented design provides base classes for rapid development of own components.

Reverberation effects as observed by room microphones severely degrade the performance of automatic speech recognition systems. We investigate the use of dereverberation by spectral subtraction as proposed by Lebart and Boucher and introduce a simple approach to estimate the required decay parameter by clapping hands. Experiments on small vocabulary continuous speech recognition task on read speech show that using the calibrated dereverberation improves WER from 73.2 to 54.7 for the best microphone. In combination with system adaptation, the WER could be reduced to 28.2, which is only a 16% relative loss of performance comparison to using a headset instead of a room microphone.

Learning and programming environments used in computer science education give feedback to the users by system messages. These are triggered by programming errors and give only technical hints without regard to the learners' problem solving process. To adapt the messages not only to the factual but also to the procedural knowledge of the learners, their problem solving strategies have to be identified automatically and in process. This article describes a way to achieve this with the help of pattern recognition methods. Using data from a study with 65 learners aged 12 to 13 using a learning environment for programming, a classification system based on hidden Markov models is trained and integrated in the very same environment. We discuss findings in that data and the performance of the automatic online identification, and present first results using the developed software in class.

The CALO Meeting Assistant (MA) provides for distributed meeting capture, annotation, automatic transcription and semantic analysis of multiparty meetings, and is part of the larger CALO personal assistant system. This paper presents the CALO-MA architecture and its speech recognition and understanding components, which include real-time and offline speech transcription, dialog act segmentation and tagging, topic identification and segmentation, question-answer pair identification, action item recognition, decision extraction, and summarization.

We introduce a model for extractive meeting summarization based on the hypothesis that utterances convey bits of information, or concepts. Using keyphrases as concepts weighted by frequency, and an integer linear program to determine the best set of utterances, that is, covering as many concepts as possible while satisfying a length constraint, we achieve ROUGE scores at least as good as a ROUGE-based oracle derived from human summaries. This brings us to a critical discussion of ROUGE and the future of extractive meeting summarization.

This paper presents an unsupervised, graph based approach for extractive summarization of meetings. Graph based methods such as TextRank have been used for sentence extraction from news articles. These methods model text as a graph with sentences as nodes and edges based on word overlap. A sentence node is then ranked according to its similarity with other nodes. The spontaneous speech in meetings leads to incomplete, informed sentences with high redundancy and calls for additional measures to extract relevant sentences. We propose an extension of the TextRank algorithm that clusters the meeting utterances and uses these clusters to construct the graph. We evaluate this method on the AM I meeting corpus and show a significant improvement over TextRank and other baseline methods.

Despite considerable work in automatic meeting summarization over the last few years, comparing results remains difficult due to varied task conditions and evaluations. To address this issue, we present a method for determining the best possible extractive summary given an evaluation metric like ROUGE. Our oracle system is based on a knapsack-packing framework, and though NP-Hard, can be solved nearly optimally by a genetic algorithm. To frame new research results in a meaningful context, we suggest presenting our oracle results alongside two simple baselines. We show oracle and baseline results for a variety of evaluation scenarios that have recently appeared in this field.

Rooted in multi-document summarization, maximum marginal relevance (MMR) is a widely used algorithm for meeting summarization (MS). A major problem in extractive MS using MMR is finding a proper query: the centroid based query which is commonly used in the absence of a manually specified query, can not significantly outperform a simple baseline system. We introduce a simple yet robust algorithm to automatically extract keyphrases (KP) from a meeting which can then be used as a query in the MMR algorithm. We show that the KP based system significantly outperforms both baseline and centroid based systems. As human refined KPs show even better summarization performance, we outline how to integrate the KP approach into a graphical user interface allowing interactive summarization to match the user's needs in terms of summary length and topic focus.

The CALO Meeting Assistant provides for distributed meeting capture, annotation, automatic transcription and semantic analysis of multiparty meetings, and is part of the larger CALO personal assistant system. This paper summarizes the CALO-MA architecture and its speech recognition and understanding components, which include real-time and offline speech transcription, dialog act segmentation and tagging, question-answer pair identification, action item recognition, decision extraction, and summarization.

For many aspects of speech therapy an objective evaluation of the intelligibility of a patient's speech is needed. We investigate the evaluation of the intelligibility of speech by means of automatic speech recognition. Previous studies have shown that measures like word accuracy are consistent with human experts' ratings. To ease the patient's burden, it is highly desirable to conduct the assessment via phone. However, the telephone channel influences the quality of the speech signal which negatively affects the results. To reduce inaccuracies, we propose a combination of two speech recognizers. Experiments on two sets of pathological speech show that the combination results in consistent improvements in the correlation between the automatic evaluation and the ratings by human experts. Furthermore, the approach leads to reductions of 10% and 25% of the maximum error of the intelligibility measure.