Automated Spoken Language Assessment

oollim is a locally executable system for evaluating L2 English pronunciation accuracy, fluency, and prosody on a consumer computer. The system combines pretrained XLS-R, XEUS, and PhoneticXeus representations with reference-text features and ridge regressions trained on SpeechOcean762, the Korean Learners' L2 Speech Corpus (English), J-AESOP, and their combinations. On 305 independently held-out recordings assessed by three expert raters, several ridge configurations substantially outperform archived Azure API scores obtained in January 2026 in both Spearman correlation and quadratic weighted kappa across all three constructs. These results support expert speech evaluation on consumer hardware using pretrained representations and construct-specific regressions, without neural-model fine-tuning or calibration on the private evaluation data.

Mispronunciation detection and diagnosis (MDD) provides phone-level feedback to language learners, but extending it to new languages typically requires collecting and annotating learner speech for each target language. This repeated annotation effort is costly and limits the language coverage of pronunciation feedback. We study whether MDD capabilities learned from annotated learner speech can be applied to languages unseen during MDD training when canonical pronunciation references are available. Our multilingual LLM-based model, X-MDD, uses canonical speech and phones as references to predict the phone sequence actually produced by a learner. It incorporates in-context demonstrations for inference-time adaptation and is further optimized with group relative policy optimization (GRPO) using detection F1 and phone error rate. We further analyze how canonical inputs affect prediction and find that (1) canonical speech and phones jointly contribute to actual-phone prediction, and (2) conflicting speech evidence reduces the tendency to copy canonical phones. Trained only on Mandarin and English, X-MDD achieves a detection F1 of 38.07 on Arabic Qur'anic recitation, compared with 44.14 for a supervised challenge baseline trained on Arabic. It also achieves F1 scores of 78.68 and 58.78 on Mandarin and English, respectively, demonstrating effective multilingual MDD within a single model.

Prosodic stress is a crucial aspect of automatic pronunciation assessment (APA), encompassing both sentence stress detection (SSD) and word stress detection (WSD). SSD highlights semantically salient words that shape discourse meaning, while WSD identifies the primary stressed syllable within each word to ensure lexical clarity. However, most prior work treats SSD and WSD as independent tasks, overlooking their shared reliance on prosodic cues such as pitch, duration, and intensity. To address this gap, we propose an effective SSD approach combining SSD with auxiliary WSD via a novel modeling paradigm. In addition, we introduce a word-span stress regularizer (WSR) that concentrates token-level SSD probabilities within each stressed word span. Experiments on the TinyStress-15K benchmark show that the proposed method outperforms strong baselines, with the complete configuration achieving the best SSD result.

The development of automated speaking assessment (ASA) is limited by the scarcity of public datasets, with most existing work relying on read-aloud speech, which limits applicability to real-world communication scenarios. In this work, we introduce OpenEnded, a corpus of English practice speech from Mandarin speakers in open-response tasks. Unlike prior open-response datasets that provide only holistic proficiency scores, OpenEnded offers utterance-level assessments of accuracy, fluency, and prosody. Approximately 10,000 utterances are collected and annotated using a hybrid framework: 1,000 are manually labeled via multi-rater scoring with discrepancy resolution to form a high-quality test set, while the remaining are pseudo-labeled by an audio language model (ALM) for training and development sets. We evaluate ALMs and existing ASA models on the OpenEnded test set and introduce VoxPA as an additional baseline. Results show that ALM-generated pseudo-labels improve training over original ALM scoring, while VoxPA achieves the best performance among all baselines.

Pronunciation assessment requires acoustic evidence that is temporally precise, diagnostically meaningful, and faithful to the learner's actual production. However, existing acoustic models often struggle to provide recognition and segmentation evidence simultaneously. CTC-based phone recognizers can predict phone sequences flexibly, but their sparse and peaky posteriors often miss phone boundaries and fine-grained pronunciation cues. In contrast, text-dependent forced aligners provide reliable temporal information when transcripts are available, but are not directly applicable to reference-free pronunciation analysis. In this work, we propose a topology-aware frame-wise acoustic model that learns dense ordered state posteriors within each phone. The key idea is to recover phone-internal state structure in a neural acoustic model by combining ordered subphonetic states with optimal temporal transport classification (OTTC). This combination encourages dense monotonic frame-level state discrimination while preserving phone recognition ability. Experiments on read, spontaneous, and L2 speech show improved segmentation over neural baselines with competitive recognition performance. Downstream evaluations further show gains in mispronunciation detection and automatic pronunciation assessment. Probing analysis suggests that the learned states capture phoneme-dependent acoustic structure rather than arbitrary frame-level distributions.

Automatic Pronunciation Assessment (APA) systems have achieved strong performance with transformer-based models and self-supervised speech representations. However, most methods rely only on acoustic signals and overlook temporal synchronization between speech and articulatory movements, limiting diagnostic feedback on timing mismatches important for L2 pronunciation training. We propose an interpretable audio-visual synchronization framework that explicitly models speech-lip temporal alignment through feature encoding, cross-attention fusion, lag estimation, stability quantification, and visualization. The framework introduces frame-level lag trajectories and a Lag Stability Index (LSI) to quantify synchronization robustness. We also interviewed 30 participants, including 10 instructors and 20 students with diverse first-language backgrounds, to assess its effectiveness. By transforming implicit alignment into interpretable representations, the framework connects automatic scoring with actionable Computer-Aided Pronunciation Training feedback. Datasets and supplemental materials are available at https://www.robots.ox.ac.uk/~vgg/data/lip_reading/.

Research on automatic speaking assessment (ASA) has increasingly adopted multimodal speech large language models to assess learners' speaking performance. However, existing studies provide limited analysis of how acoustic and content information contribute to predictions and how stable the resulting performance is. We propose CASA, a simpler architecture combining Whisper-medium and Qwen3.5-2B that achieves state-of-the-art performance while providing a more interpretable separation between speech delivery and content. On the Speak & Improve Corpus 2025, CASA achieves a root mean square error (RMSE) of 0.358, improving on the previous best RMSE while using approximately half the estimated inference parameters. The general-purpose architecture is designed for adaptation to other ASA corpora without structural changes and relies on three handcrafted fluency features. Through ablations and repeated runs, we analyze the individual and complementary contributions of acoustic and content information, examine performance variability, and demonstrate the potential of large language model reasoning for training-free content validation.

Automatic speaking assessment systems are increasingly deployed in high-stakes settings to mark second language (L2) learners' speaking tests, making it critical to show that their scores depend on speaking proficiency rather than irrelevant speaker attributes such as first language (L1) or age. Transformer-based foundation models have improved the accuracy of these L2 speaking graders, but their black-box representations make fairness and interpretability analysis more difficult. Building on prior work that used Concept Activation Vectors (CAVs) to detect bias towards unwanted attributes (`concepts') in feature-based graders, we extend CAV-based analysis to two neural speaking assessment systems: a text-based BERT grader and a speech-and-text multimodal grader based on Whisper. CAVs represent human-interpretable concepts as directions in a model's activation space, allowing us to distinguish between whether a concept is encoded in a model's internal representations and whether it influences the predicted score, the latter quantified using a gradient-based sensitivity metric. Since CAVs rely on linear separability, which is less likely in complex neural embedding spaces, we also investigate whether sparse autoencoders (SAEs) provide cleaner concept directions by learning CAVs in a sparse latent space and mapping them back to activation space. Our analysis shows that concept recoverability depends strongly on the representation and architecture being probed, rather than on the concept alone. Sensitivity to concepts is also architecture-dependent. SAEs make concepts more linearly recoverable, but attenuate the original activation-space sensitivity, especially in low-dimensional layers. These findings highlight the need to distinguish concept recoverability from concept influence when auditing bias in speaking assessment systems.

Increasingly, speech and language processing tasks take either audio or text directly rather than extracting features from these as the input to the classifier or regressor. Often these systems make use of complex, for example transformer-based, processes that have the ability to derive highly non-linear mappings between the input and the output. Unfortunately these systems can also learn ''shortcuts'' where the classifier is overly reliant on particular aspects of the input to yield the output. For the task of language proficiency assessment, this over-reliance can enable learners to increase their score by exploiting the shortcut rather than improving their ability. This paper introduces a novel training criterion that is able to reduce the classifier's reliance on shortcuts, thus for example limiting this option for malpractice in language assessment. This process is illustrated on two forms of assessment system, one based on the audio the other on the speech recognition text. The results show that, for both systems, there is higher correlations with features that could be exploited for malpractice than expected from the human reference, indicating an over-reliance on these features. By introducing the modified training criterion, this correlation can be reduced to be closer to the reference correlation.

Training automated pronunciation assessment often relies on labeled learner errors or non-native corpora that are costly to collect. We propose a lightweight framework trained only on native speech resources, operating unsupervised or lightly calibrated with a small set of scored utterances. At inference, learner speech is discretized with an SSL encoder and a K-means codebook. A token language model trained on native sequences computes surprisal where higher surprisal indicates phonotactic deviation. We add a transcript-guided Text2DUnit--DTW module that predicts native token sequences from reference text and aligns them to acoustic tokens to derive error-sensitive features. Surprisal and alignment features are fused via simple regression. On SpeechOcean762, PCC improves from 0.60 to 0.66 with transcript guidance, near supervised baselines. Cross-dataset evaluation on L2-ARCTIC shows consistent gains.

Automated L2 speech assessment can assign proficiency labels, but often lacks interpretability. We propose a rubric-guided SpeechLLM for multi-aspect, multi-granular assessment, trained with a hybrid objective combining supervised fine-tuning and Bounded Direct Preference Optimization. The model jointly predicts ordinal labels at the sentence-level (accuracy, fluency, prosody), word/phoneme-level accuracy, and generates a natural-language rationale in the same response. On SpeechOcean762, our approach matches or outperforms single-granularity models while remaining competitive with prior approaches. We analyze rationale reliability along two axes: self-consistency with model predictions and alignment with ground-truth labels, using sentiment consistency (plausibility) and mention-based agreement (faithfulness). Rationales are plausible at the sentence level, but faithfulness degrades at the word/phoneme level: references are sparse and weakly aligned with token-level labels.

We present the findings of the second edition of the IQRA Interspeech Challenge, a challenge on automatic Mispronunciation Detection and Diagnosis (MDD) for Modern Standard Arabic (MSA). Building on the previous edition, this iteration introduces Iqra_Extra_IS26, a new dataset of authentic human mispronounced speech, complementing the existing training and evaluation resources. Submitted systems employed a diverse range of approaches, spanning CTC-based self-supervised learning models, two-stage fine-tuning strategies, and using large audio-language models. Compared to the first edition, we observe a substantial jump of 0.28 in F1-score, attributable both to novel architectures and modeling strategies proposed by participants and to the additional authentic mispronunciation data made available. These results demonstrate the growing maturity of Arabic MDD research and establish a stronger foundation for future work in Arabic pronunciation assessment.

In speech evaluation, an Automatic Speech Recognition (ASR) model often computes time boundaries and phoneme posteriors for input features. However, limited data for ASR training hinders expansion of speech evaluation to low-resource languages. Open-source weakly-supervised models are capable of ASR over many languages, but they are frame-asynchronous and not phonemic, hindering feature extraction for speech evaluation. This paper proposes to overcome incompatibilities for feature extraction with weakly-supervised models, easing expansion of speech evaluation to low-resource languages. Phoneme posteriors are computed by mapping ASR hypotheses to a phoneme confusion network. Word instead of phoneme-level speaking rate and duration are used. Phoneme and frame-level features are combined using a cross-attention architecture, obviating phoneme time alignment. This performs comparably with standard frame-synchronous features on English speechocean762 and low-resource Tamil datasets.

A recent line of research on automated speaking assessment (ASA) has benefited from self-supervised learning (SSL) representations, which capture rich acoustic and linguistic patterns in non-native speech without underlying assumptions of feature curation. However, speech-based SSL models capture acoustic-related traits but overlook linguistic content, while text-based SSL models rely on ASR output and fail to encode prosodic nuances. Moreover, most prior arts treat proficiency levels as nominal classes, ignoring their ordinal structure and non-uniform intervals between proficiency labels. To address these limitations, we propose an effective ASA approach combining SSL with handcrafted indicator features via a novel modeling paradigm. We further introduce a multi-margin ordinal loss that jointly models both the score ordinality and non-uniform intervals of proficiency labels. Extensive experiments on the TEEMI corpus show that our method consistently outperforms strong baselines and generalizes well to unseen prompts.

A recent line of research on spoken language assessment (SLA) employs neural models such as BERT and wav2vec 2.0 (W2V) to evaluate speaking proficiency across linguistic and acoustic modalities. Although both models effectively capture features relevant to oral competence, each exhibits modality-specific limitations. BERT-based methods rely on ASR transcripts, which often fail to capture prosodic and phonetic cues for SLA. In contrast, W2V-based methods excel at modeling acoustic features but lack semantic interpretability. To overcome these limitations, we propose a system that integrates W2V with Phi-4 multimodal large language model (MLLM) through a score fusion strategy. The proposed system achieves a root mean square error (RMSE) of 0.375 on the official test set of the Speak & Improve Challenge 2025, securing second place in the competition. For comparison, the RMSEs of the top-ranked, third-ranked, and official baseline systems are 0.364, 0.384, and 0.444, respectively.