Multilingual Gloss-free Sign Language Translation: Towards Building a Sign Language Foundation Model

Sign Language Translation (SLT) aims to convert sign language (SL) videos into spoken language text, thereby bridging the communication gap between the sign and the spoken community. While most existing works focus on translating a single sign language into a single spoken language (one-to-one SLT), leveraging multilingual resources could mitigate low-resource issues and enhance accessibility. However, multilingual SLT (MLSLT) remains unexplored due to language conflicts and alignment difficulties across SLs and spoken languages. To address these challenges, we propose a multilingual gloss-free model with dual CTC objectives for token-level SL identification and spoken text generation. Our model supports 10 SLs and handles one-to-one, many-to-one, and many-to-many SLT tasks, achieving competitive performance compared to state-of-the-art methods on three widely adopted benchmarks: multilingual SP-10, PHOENIX14T, and CSL-Daily.

Dynamic Context-Aware Streaming Pretrained Language Model For Inverse Text Normalization

Inverse Text Normalization (ITN) is crucial for converting spoken Automatic Speech Recognition (ASR) outputs into well-formatted written text, enhancing both readability and usability. Despite its importance, the integration of streaming ITN within streaming ASR remains largely unexplored due to challenges in accuracy, efficiency, and adaptability, particularly in low-resource and limited-context scenarios. In this paper, we introduce a streaming pretrained language model for ITN, leveraging pretrained linguistic representations for improved robustness. To address streaming constraints, we propose Dynamic Context-Aware during training and inference, enabling adaptive chunk size adjustments and the integration of right-context information. Experimental results demonstrate that our method achieves accuracy comparable to non-streaming ITN and surpasses existing streaming ITN models on a Vietnamese dataset, all while maintaining low latency, ensuring seamless integration into ASR systems.

Voice Conversion Improves Cross-Domain Robustness for Spoken Arabic Dialect Identification

Arabic dialect identification (ADI) systems are essential for large-scale data collection pipelines that enable the development of inclusive speech technologies for Arabic language varieties. However, the reliability of current ADI systems is limited by poor generalization to out-of-domain speech. In this paper, we present an effective approach based on voice conversion for training ADI models that achieves state-of-the-art performance and significantly improves robustness in cross-domain scenarios. Evaluated on a newly collected real-world test set spanning four different domains, our approach yields consistent improvements of up to +34.1% in accuracy across domains. Furthermore, we present an analysis of our approach and demonstrate that voice conversion helps mitigate the speaker bias in the ADI dataset. We release our robust ADI model and cross-domain evaluation dataset to support the development of inclusive speech technologies for Arabic.

Spoken Language Modeling with Duration-Penalized Self-Supervised Units

Spoken language models (SLMs) operate on acoustic units obtained by discretizing self-supervised speech representations. Although the characteristics of these units directly affect performance, the interaction between codebook size and unit coarseness (i.e., duration) remains unexplored. We investigate SLM performance as we vary codebook size and unit coarseness using the simple duration-penalized dynamic programming (DPDP) method. New analyses are performed across different linguistic levels. At the phone and word levels, coarseness provides little benefit, as long as the codebook size is chosen appropriately. However, when producing whole sentences in a resynthesis task, SLMs perform better with coarser units. In lexical and syntactic language modeling tasks, coarser units also give higher accuracies at lower bitrates. We therefore show that coarser units aren't always better, but that DPDP is a simple and efficient way to obtain coarser units for the tasks where they are beneficial.

LLM-Synth4KWS: Scalable Automatic Generation and Synthesis of Confusable Data for Custom Keyword Spotting

Custom keyword spotting (KWS) allows detecting user-defined spoken keywords from streaming audio. This is achieved by comparing the embeddings from voice enrollments and input audio. State-of-the-art custom KWS models are typically trained contrastively using utterances whose keywords are randomly sampled from training dataset. These KWS models often struggle with confusing keywords, such as "blue" versus "glue". This paper introduces an effective way to augment the training with confusable utterances where keywords are generated and grouped from large language models (LLMs), and speech signals are synthesized with diverse speaking styles from text-to-speech (TTS) engines. To better measure user experience on confusable KWS, we define a new northstar metric using the average area under DET curve from confusable groups (c-AUC). Featuring high scalability and zero labor cost, the proposed method improves AUC by 3.7% and c-AUC by 11.3% on the Speech Commands testing set.

Spoken question answering for visual queries

Question answering (QA) systems are designed to answer natural language questions. Visual QA (VQA) and Spoken QA (SQA) systems extend the textual QA system to accept visual and spoken input respectively. This work aims to create a system that enables user interaction through both speech and images. That is achieved through the fusion of text, speech, and image modalities to tackle the task of spoken VQA (SVQA). The resulting multi-modal model has textual, visual, and spoken inputs and can answer spoken questions on images. Training and evaluating SVQA models requires a dataset for all three modalities, but no such dataset currently exists. We address this problem by synthesizing VQA datasets using two zero-shot TTS models. Our initial findings indicate that a model trained only with synthesized speech nearly reaches the performance of the upper-bounding model trained on textual QAs. In addition, we show that the choice of the TTS model has a minor impact on accuracy.

In-context Language Learning for Endangered Languages in Speech Recognition

With approximately 7,000 languages spoken worldwide, current large language models (LLMs) support only a small subset. Prior research indicates LLMs can learn new languages for certain tasks without supervised data. We extend this investigation to speech recognition, investigating whether LLMs can learn unseen, low-resource languages through in-context learning (ICL). With experiments on four diverse endangered languages that LLMs have not been trained on, we find that providing more relevant text samples enhances performance in both language modelling and Automatic Speech Recognition (ASR) tasks. Furthermore, we show that the probability-based approach outperforms the traditional instruction-based approach in language learning. Lastly, we show ICL enables LLMs to achieve ASR performance that is comparable to or even surpasses dedicated language models trained specifically for these languages, while preserving the original capabilities of the LLMs.

Counting trees: A treebank-driven exploration of syntactic variation in speech and writing across languages

This paper presents a novel treebank-driven approach to comparing syntactic structures in speech and writing using dependency-parsed corpora. Adopting a fully inductive, bottom-up method, we define syntactic structures as delexicalized dependency (sub)trees and extract them from spoken and written Universal Dependencies (UD) treebanks in two syntactically distinct languages, English and Slovenian. For each corpus, we analyze the size, diversity, and distribution of syntactic inventories, their overlap across modalities, and the structures most characteristic of speech. Results show that, across both languages, spoken corpora contain fewer and less diverse syntactic structures than their written counterparts, with consistent cross-linguistic preferences for certain structural types across modalities. Strikingly, the overlap between spoken and written syntactic inventories is very limited: most structures attested in speech do not occur in writing, pointing to modality-specific preferences in syntactic organization that reflect the distinct demands of real-time interaction and elaborated writing. This contrast is further supported by a keyness analysis of the most frequent speech-specific structures, which highlights patterns associated with interactivity, context-grounding, and economy of expression. We argue that this scalable, language-independent framework offers a useful general method for systematically studying syntactic variation across corpora, laying the groundwork for more comprehensive data-driven theories of grammar in use.

Large Language Models for Depression Recognition in Spoken Language Integrating Psychological Knowledge

Depression is a growing concern gaining attention in both public discourse and AI research. While deep neural networks (DNNs) have been used for recognition, they still lack real-world effectiveness. Large language models (LLMs) show strong potential but require domain-specific fine-tuning and struggle with non-textual cues. Since depression is often expressed through vocal tone and behaviour rather than explicit text, relying on language alone is insufficient. Diagnostic accuracy also suffers without incorporating psychological expertise. To address these limitations, we present, to the best of our knowledge, the first application of LLMs to multimodal depression detection using the DAIC-WOZ dataset. We extract the audio features using the pre-trained model Wav2Vec, and mapped it to text-based LLMs for further processing. We also propose a novel strategy for incorporating psychological knowledge into LLMs to enhance diagnostic performance, specifically using a question and answer set to grant authorised knowledge to LLMs. Our approach yields a notable improvement in both Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) compared to a base score proposed by the related original paper. The codes are available at https://github.com/myxp-lyp/Depression-detection.git

StressTest: Can YOUR Speech LM Handle the Stress?

Sentence stress refers to emphasis, placed on specific words within a spoken utterance to highlight or contrast an idea, or to introduce new information. It is often used to imply an underlying intention that is not explicitly stated. Recent advances in speech-aware language models (SLMs) have enabled direct processing of audio, allowing models to bypass transcription and access the full richness of the speech signal and perform audio reasoning tasks such as spoken question answering. Despite the crucial role of sentence stress in shaping meaning and speaker intent, it remains largely overlooked in evaluation and development of such models. In this work, we address this gap by introducing StressTest, a benchmark specifically designed to evaluate a model's ability to distinguish between interpretations of spoken sentences based on the stress pattern. We assess the performance of several leading SLMs and find that, despite their overall capabilities, they perform poorly on such tasks. To overcome this limitation, we propose a novel synthetic data generation pipeline, and create Stress17k, a training set that simulates change of meaning implied by stress variation. Then, we empirically show that optimizing models with this synthetic dataset aligns well with real-world recordings and enables effective finetuning of SLMs. Results suggest, that our finetuned model, StresSLM, significantly outperforms existing models on both sentence stress reasoning and detection tasks. Code, models, data, and audio samples - pages.cs.huji.ac.il/adiyoss-lab/stresstest.