Automatic dialect classification is commonly treated as a supervised task with a primary focus on overall accuracy. In this paper, we argue that classification errors and model uncertainty provide valuable insights into dialectal structure and variation. We analyze a speech-based dialect classification model trained on German dialect data from three generations and evaluated across 250 speaker-disjoint splits (median weighted F1=0.42). A systematic confusion analysis shows that misclassifications are largely explained by speaker diversity, dialectal similarity, geographical proximity, and speaker self-assessment. Among these factors, the number of speakers per dialect has the strongest impact on performance, while frequent confusions between closely related dialects reflect inherent linguistic similarity rather than model limitations. Generational analyses further indicate that younger speakers exhibit reduced dialectal distinctiveness, although core dialectal features remain shared across generations. By explicitly modeling classification uncertainty, the proposed approach enables the analysis of dialect transition areas and gradient dialect boundaries. Overall, this work demonstrates that automatic dialect classification can serve not only as a predictive task but also as a tool for dialectological analysis.

Due to the scarcity of labeled dialectal speech, audio dialect classification is a challenging task for most languages, including Swiss German. In this work, we explore the ability of large language models (LLMs) as agents in understanding the dialects and whether they can show comparable performance to models such as HuBERT in dialect classification. In addition, we provide an LLM baseline and a human linguist one. Our approach uses phonetic transcriptions produced by ASR systems and combines them with linguistic resources such as dialect feature maps, vowel history, and rules. Our findings indicate that, when linguistic information is provided, the LLM predictions improve. The human baseline shows that automatically generated transcriptions can be beneficial for such classifications, but also present opportunities for improvement.

This study evaluates whether Retrieval-based Voice Conversion (RVC) can be used to normalize speaker-specific variability while preserving dialect-relevant acoustic cues, and what the response of human and machine systems to this manipulation reveals about the architecture of dialect recognition. In two perception experiments, speech samples from nine German dialect regions were presented either in their original form or after conversion to a single target speaker. We compared overall accuracy, confusion structures, item-level response distributions, and the interaction between listener origin and target dialect across conditions. Human classification remained stable under voice conversion. Accuracy did not differ between conditions, confusion matrices were highly correlated, and item-level divergences were minimal. The interaction between listener origin and target dialect—reflecting systematic regional bias—remained invariant. These findings indicate that RVC does not distort perceptually relevant dialectal cues and that human dialect recognition is robust to speaker normalization. In contrast, we evaluated a deep learning model under matched conditions: model accuracy improved significantly under RVC, while human performance remained unchanged. This dissociation reframes RVC as an experimental probe for investigating the divergence between human and machine speech processing, suggesting that this divergence is rooted in fundamentally different representational architectures.

We investigate lightweight and easily applicable data augmentation techniques for dialectal audio classification. We evaluate four main methods, namely shifting pitch, interval removal, background noise insertion and interval swap as well as several subvariants on recordings from 20 German dialects. Each main method is tested across multiple hyperparameter combinations, inlcuding augmentation length, coverage ratio and number of augmentations per original sample. Our results show that frequency-based techniques, particularly frequency masking, consistently yield performance improvements, while others such as time masking or speaker-based insertion can negatively affect the results. Our comparative analysis identifies which augmentations are most effective under realistic conditions, offering simple and efficient strategies to improve dialectal speech classification.

This paper examines the effect of preprocessing techniques on spoken dialect classification using raw audio data. We focus on modifying Root Mean Square (RMS) amplitude, DC-offset, articulation rate (AR), pitch, and Harmonics-to-Noise Ratio (HNR) to assess their impact on model performance. Our analysis determines whether these features are important, irrelevant, or misleading for the classification task. To evaluate these effects, we use a pipeline that tests the significance of each acoustic feature through distortion and normalization techniques. While preprocessing did not directly improve classification accuracy, our findings reveal three key insights: deep learning models for dialect classification are generally robust to variations in the tested audio features, suggesting that normalization may not be necessary. We identify articulation rate as a critical factor, directly affecting the amount of information in audio chunks. Additionally, we demonstrate that intonation, specifically the pitch range, plays a vital role in dialect recognition.

Speaker diarization is a critical task in the field of computer science, aiming to assign timestamps and speaker labels to audio segments. The aim of these tests in this Publication is to find a pretrained speaker diarization pipeline capable of distinguishing dialectal speakers from each other and an explorer. To achieve this, three pipelines, namely Pyannote, CLEAVER and NeMo, are tested and compared, across various segmentation and parameterization strategies. The study considers multiple scenarios, such as the impact of threshold values, overlap handling, and minimum duration parameters, on classification accuracy. Additionally, this study aims to create a dataset for German dialect identification (DID) based on the findings from this research.