Mauro Cettolo

Marco Gaido, Sara Papi, Mauro Cettolo et al.

Streaming Speech-to-Text Translation (StreamST) requires producing translations concurrently with incoming speech, imposing strict latency constraints and demanding models that balance partial-information decision-making with high translation quality. Research efforts on the topic have so far relied on the SimulEval repository, which is no longer maintained and does not support systems that revise their outputs. In addition, it has been designed for simulating the processing of short segments, rather than long-form audio streams, and it does not provide an easy method to showcase systems in a demo. As a solution, we introduce simulstream, the first open-source framework dedicated to unified evaluation and demonstration of StreamST systems. Designed for long-form speech processing, it supports not only incremental decoding approaches, but also re-translation methods, enabling for their comparison within the same framework both in terms of quality and latency. In addition, it also offers an interactive web interface to demo any system built within the tool.

Sara Papi, Marco Gaido, Alina Karakanta et al.

Automatic subtitling is the task of automatically translating the speech of audiovisual content into short pieces of timed text, i.e. subtitles and their corresponding timestamps. The generated subtitles need to conform to space and time requirements, while being synchronised with the speech and segmented in a way that facilitates comprehension. Given its considerable complexity, the task has so far been addressed through a pipeline of components that separately deal with transcribing, translating, and segmenting text into subtitles, as well as predicting timestamps. In this paper, we propose the first direct ST model for automatic subtitling that generates subtitles in the target language along with their timestamps with a single model. Our experiments on 7 language pairs show that our approach outperforms a cascade system in the same data condition, also being competitive with production tools on both in-domain and newly-released out-domain benchmarks covering new scenarios.

Alina Karakanta, François Buet, Mauro Cettolo et al.

Subtitles appear on screen as short pieces of text, segmented based on formal constraints (length) and syntactic/semantic criteria. Subtitle segmentation can be evaluated with sequence segmentation metrics against a human reference. However, standard segmentation metrics cannot be applied when systems generate outputs different than the reference, e.g. with end-to-end subtitling systems. In this paper, we study ways to conduct reference-based evaluations of segmentation accuracy irrespective of the textual content. We first conduct a systematic analysis of existing metrics for evaluating subtitle segmentation. We then introduce $Sigma$, a new Subtitle Segmentation Score derived from an approximate upper-bound of BLEU on segmentation boundaries, which allows us to disentangle the effect of good segmentation from text quality. To compare $Sigma$ with existing metrics, we further propose a boundary projection method from imperfect hypotheses to the true reference. Results show that all metrics are able to reward high quality output but for similar outputs system ranking depends on each metric's sensitivity to error type. Our thorough analyses suggest $Sigma$ is a promising segmentation candidate but its reliability over other segmentation metrics remains to be validated through correlations with human judgements.

Dennis Fucci, Marco Gaido, Matteo Negri et al.

Automatic speech recognition (ASR) systems are known to be sensitive to the sociolinguistic variability of speech data, in which gender plays a crucial role. This can result in disparities in recognition accuracy between male and female speakers, primarily due to the under-representation of the latter group in the training data. While in the context of hybrid ASR models several solutions have been proposed, the gender bias issue has not been explicitly addressed in end-to-end neural architectures. To fill this gap, we propose a data augmentation technique that manipulates the fundamental frequency (f0) and formants. This technique reduces the data unbalance among genders by simulating voices of the under-represented female speakers and increases the variability within each gender group. Experiments on spontaneous English speech show that our technique yields a relative WER improvement up to 9.87% for utterances by female speakers, with larger gains for the least-represented f0 ranges.

Dennis Fucci, Marco Gaido, Sara Papi et al.

When translating words referring to the speaker, speech translation (ST) systems should not resort to default masculine generics nor rely on potentially misleading vocal traits. Rather, they should assign gender according to the speakers' preference. The existing solutions to do so, though effective, are hardly feasible in practice as they involve dedicated model re-training on gender-labeled ST data. To overcome these limitations, we propose the first inference-time solution to control speaker-related gender inflections in ST. Our approach partially replaces the (biased) internal language model (LM) implicitly learned by the ST decoder with gender-specific external LMs. Experiments on en->es/fr/it show that our solution outperforms the base models and the best training-time mitigation strategy by up to 31.0 and 1.6 points in gender accuracy, respectively, for feminine forms. The gains are even larger (up to 32.0 and 3.4) in the challenging condition where speakers' vocal traits conflict with their gender.

Malvina Nissim, Danilo Croce, Viviana Patti et al.

The rapid progress of Large Language Models (LLMs) has transformed natural language processing and broadened its impact across research and society. Yet, systematic evaluation of these models, especially for languages beyond English, remains limited. "Challenging the Abilities of LAnguage Models in ITAlian" (CALAMITA) is a large-scale collaborative benchmarking initiative for Italian, coordinated under the Italian Association for Computational Linguistics. Unlike existing efforts that focus on leaderboards, CALAMITA foregrounds methodology: it federates more than 80 contributors from academia, industry, and the public sector to design, document, and evaluate a diverse collection of tasks, covering linguistic competence, commonsense reasoning, factual consistency, fairness, summarization, translation, and code generation. Through this process, we not only assembled a benchmark of over 20 tasks and almost 100 subtasks, but also established a centralized evaluation pipeline that supports heterogeneous datasets and metrics. We report results for four open-weight LLMs, highlighting systematic strengths and weaknesses across abilities, as well as challenges in task-specific evaluation. Beyond quantitative results, CALAMITA exposes methodological lessons: the necessity of fine-grained, task-representative metrics, the importance of harmonized pipelines, and the benefits and limitations of broad community engagement. CALAMITA is conceived as a rolling benchmark, enabling continuous integration of new tasks and models. This makes it both a resource -- the most comprehensive and diverse benchmark for Italian to date -- and a framework for sustainable, community-driven evaluation. We argue that this combination offers a blueprint for other languages and communities seeking inclusive and rigorous LLM evaluation practices.

Mauro Cettolo, Marco Gaido, Matteo Negri et al.

Automatic evaluation of speech-to-text translation (ST) systems is typically performed by comparing translation hypotheses with one or more reference translations. While effective to some extent, this approach inherits the limitation of reference-based evaluation that ignores valuable information from the source input. In machine translation (MT), recent progress has shown that neural metrics incorporating the source text achieve stronger correlation with human judgments. Extending this idea to ST, however, is not trivial because the source is audio rather than text, and reliable transcripts or alignments between source and references are often unavailable. In this work, we conduct the first systematic study of source-aware metrics for ST, with a particular focus on real-world operating conditions where source transcripts are not available. We explore two complementary strategies for generating textual proxies of the input audio, automatic speech recognition (ASR) transcripts, and back-translations of the reference translation, and introduce a novel two-step cross-lingual re-segmentation algorithm to address the alignment mismatch between synthetic sources and reference translations. Our experiments, carried out on two ST benchmarks covering 79 language pairs and six ST systems with diverse architectures and performance levels, show that ASR transcripts constitute a more reliable synthetic source than back-translations when word error rate is below 20%, while back-translations always represent a computationally cheaper but still effective alternative. Furthermore, our cross-lingual re-segmentation algorithm enables robust use of source-aware MT metrics in ST evaluation, paving the way toward more accurate and principled evaluation methodologies for speech translation.

Sara Papi, Marco Gaido, Luisa Bentivogli et al.

The development of speech foundation models (SFMs) like Whisper and SeamlessM4T has significantly advanced the field of speech processing. However, their closed nature--with inaccessible training data and code--poses major reproducibility and fair evaluation challenges. While other domains have made substantial progress toward open science by developing fully transparent models trained on open-source (OS) code and data, similar efforts in speech remain limited. To fill this gap, we introduce FAMA, the first family of open science SFMs for English and Italian, trained on 150k+ hours of OS speech data. Moreover, we present a new dataset containing 16k hours of cleaned and pseudo-labeled speech for both languages. Results show that FAMA achieves competitive performance compared to existing SFMs while being up to 8 times faster. All artifacts, including code, datasets, and models, are released under OS-compliant licenses, promoting openness in speech technology research.

Ibrahim Said Ahmad, Antonios Anastasopoulos, Ondřej Bojar et al.

This paper reports on the shared tasks organized by the 21st IWSLT Conference. The shared tasks address 7 scientific challenges in spoken language translation: simultaneous and offline translation, automatic subtitling and dubbing, speech-to-speech translation, dialect and low-resource speech translation, and Indic languages. The shared tasks attracted 18 teams whose submissions are documented in 26 system papers. The growing interest towards spoken language translation is also witnessed by the constantly increasing number of shared task organizers and contributors to the overview paper, almost evenly distributed across industry and academia.

Dennis Fucci, Marco Gaido, Beatrice Savoldi et al.

Spurred by the demand for interpretable models, research on eXplainable AI for language technologies has experienced significant growth, with feature attribution methods emerging as a cornerstone of this progress. While prior work in NLP explored such methods for classification tasks and textual applications, explainability intersecting generation and speech is lagging, with existing techniques failing to account for the autoregressive nature of state-of-the-art models and to provide fine-grained, phonetically meaningful explanations. We address this gap by introducing Spectrogram Perturbation for Explainable Speech-to-text Generation (SPES), a feature attribution technique applicable to sequence generation tasks with autoregressive models. SPES provides explanations for each predicted token based on both the input spectrogram and the previously generated tokens. Extensive evaluation on speech recognition and translation demonstrates that SPES generates explanations that are faithful and plausible to humans.

Marco Gaido, Sara Papi, Matteo Negri et al.

Subtitling plays a crucial role in enhancing the accessibility of audiovisual content and encompasses three primary subtasks: translating spoken dialogue, segmenting translations into concise textual units, and estimating timestamps that govern their on-screen duration. Past attempts to automate this process rely, to varying degrees, on automatic transcripts, employed diversely for the three subtasks. In response to the acknowledged limitations associated with this reliance on transcripts, recent research has shifted towards transcription-free solutions for translation and segmentation, leaving the direct generation of timestamps as uncharted territory. To fill this gap, we introduce the first direct model capable of producing automatic subtitles, entirely eliminating any dependence on intermediate transcripts also for timestamp prediction. Experimental results, backed by manual evaluation, showcase our solution's new state-of-the-art performance across multiple language pairs and diverse conditions.

Dennis Fucci, Marco Gaido, Matteo Negri et al.

Despite significant advances in ASR, the specific acoustic cues models rely on remain unclear. Prior studies have examined such cues on a limited set of phonemes and outdated models. In this work, we apply a feature attribution technique to identify the relevant acoustic cues for a modern Conformer-based ASR system. By analyzing plosives, fricatives, and vowels, we assess how feature attributions align with their acoustic properties in the time and frequency domains, also essential for human speech perception. Our findings show that the ASR model relies on vowels' full time spans, particularly their first two formants, with greater saliency in male speech. It also better captures the spectral characteristics of sibilant fricatives than non-sibilants and prioritizes the release phase in plosives, especially burst characteristics. These insights enhance the interpretability of ASR models and highlight areas for future research to uncover potential gaps in model robustness.

Marco Gaido, Sara Papi, Luisa Bentivogli et al.

Training large-scale models presents challenges not only in terms of resource requirements but also in terms of their convergence. For this reason, the learning rate (LR) is often decreased when the size of a model is increased. Such a simple solution is not enough in the case of speech-to-text (S2T) trainings, where evolved and more complex variants of the Transformer architecture -- e.g., Conformer or Branchformer -- are used in light of their better performance. As a workaround, OWSM designed a double linear warmup of the LR, increasing it to a very small value in the first phase before updating it to a higher value in the second phase. While this solution worked well in practice, it was not compared with alternative solutions, nor was the impact on the final performance of different LR warmup schedules studied. This paper fills this gap, revealing that i) large-scale S2T trainings demand a sub-exponential LR warmup, and ii) a higher LR in the warmup phase accelerates initial convergence, but it does not boost final performance.

Luisa Bentivogli, Mauro Cettolo, Marco Gaido et al.

Five years after the first published proofs of concept, direct approaches to speech translation (ST) are now competing with traditional cascade solutions. In light of this steady progress, can we claim that the performance gap between the two is closed? Starting from this question, we present a systematic comparison between state-of-the-art systems representative of the two paradigms. Focusing on three language directions (English-German/Italian/Spanish), we conduct automatic and manual evaluations, exploiting high-quality professional post-edits and annotations. Our multi-faceted analysis on one of the few publicly available ST benchmarks attests for the first time that: i) the gap between the two paradigms is now closed, and ii) the subtle differences observed in their behavior are not sufficient for humans neither to distinguish them nor to prefer one over the other.

Marco Gaido, Matteo Negri, Mauro Cettolo et al.

The audio segmentation mismatch between training data and those seen at run-time is a major problem in direct speech translation. Indeed, while systems are usually trained on manually segmented corpora, in real use cases they are often presented with continuous audio requiring automatic (and sub-optimal) segmentation. After comparing existing techniques (VAD-based, fixed-length and hybrid segmentation methods), in this paper we propose enhanced hybrid solutions to produce better results without sacrificing latency. Through experiments on different domains and language pairs, we show that our methods outperform all the other techniques, reducing by at least 30% the gap between the traditional VAD-based approach and optimal manual segmentation.

Marco Gaido, Mauro Cettolo, Matteo Negri et al.

Previous studies demonstrated that a dynamic phone-informed compression of the input audio is beneficial for speech translation (ST). However, they required a dedicated model for phone recognition and did not test this solution for direct ST, in which a single model translates the input audio into the target language without intermediate representations. In this work, we propose the first method able to perform a dynamic compression of the input indirect ST models. In particular, we exploit the Connectionist Temporal Classification (CTC) to compress the input sequence according to its phonetic characteristics. Our experiments demonstrate that our solution brings a 1.3-1.5 BLEU improvement over a strong baseline on two language pairs (English-Italian and English-German), contextually reducing the memory footprint by more than 10%.

Marco Gaido, Mattia Antonino Di Gangi, Matteo Negri et al.

Direct speech-to-text translation (ST) models are usually trained on corpora segmented at sentence level, but at inference time they are commonly fed with audio split by a voice activity detector (VAD). Since VAD segmentation is not syntax-informed, the resulting segments do not necessarily correspond to well-formed sentences uttered by the speaker but, most likely, to fragments of one or more sentences. This segmentation mismatch degrades considerably the quality of ST models' output. So far, researchers have focused on improving audio segmentation towards producing sentence-like splits. In this paper, instead, we address the issue in the model, making it more robust to a different, potentially sub-optimal segmentation. To this aim, we train our models on randomly segmented data and compare two approaches: fine-tuning and adding the previous segment as context. We show that our context-aware solution is more robust to VAD-segmented input, outperforming a strong base model and the fine-tuning on different VAD segmentations of an English-German test set by up to 4.25 BLEU points.

Liane Guillou, Christian Hardmeier, Preslav Nakov et al.

We describe the design, the evaluation setup, and the results of the 2016 WMT shared task on cross-lingual pronoun prediction. This is a classification task in which participants are asked to provide predictions on what pronoun class label should replace a placeholder value in the target-language text, provided in lemmatised and PoS-tagged form. We provided four subtasks, for the English-French and English-German language pairs, in both directions. Eleven teams participated in the shared task; nine for the English-French subtask, five for French-English, nine for English-German, and six for German-English. Most of the submissions outperformed two strong language-model based baseline systems, with systems using deep recurrent neural networks outperforming those using other architectures for most language pairs.

Surafel M. Lakew, Mauro Cettolo, Marcello Federico

Recently, neural machine translation (NMT) has been extended to multilinguality, that is to handle more than one translation direction with a single system. Multilingual NMT showed competitive performance against pure bilingual systems. Notably, in low-resource settings, it proved to work effectively and efficiently, thanks to shared representation space that is forced across languages and induces a sort of transfer-learning. Furthermore, multilingual NMT enables so-called zero-shot inference across language pairs never seen at training time. Despite the increasing interest in this framework, an in-depth analysis of what a multilingual NMT model is capable of and what it is not is still missing. Motivated by this, our work (i) provides a quantitative and comparative analysis of the translations produced by bilingual, multilingual and zero-shot systems; (ii) investigates the translation quality of two of the currently dominant neural architectures in MT, which are the Recurrent and the Transformer ones; and (iii) quantitatively explores how the closeness between languages influences the zero-shot translation. Our analysis leverages multiple professional post-edits of automatic translations by several different systems and focuses both on automatic standard metrics (BLEU and TER) and on widely used error categories, which are lexical, morphology, and word order errors.

Mauro Cettolo, Mara Chinea Rios, Roldano Cattoni

In this paper, we report on domain clustering in the ambit of an adaptive MT architecture. A standard bottom-up hierarchical clustering algorithm has been instantiated with five different distances, which have been compared, on an MT benchmark built on 40 commercial domains, in terms of dendrograms, intrinsic and extrinsic evaluations. The main outcome is that the most expensive distance is also the only one able to allow the MT engine to guarantee good performance even with few, but highly populated clusters of domains.

Mauro Cettolo

We describe an Arabic-Hebrew parallel corpus of TED talks built upon WIT3, the Web inventory that repurposes the original content of the TED website in a way which is more convenient for MT researchers. The benchmark consists of about 2,000 talks, whose subtitles in Arabic and Hebrew have been accurately aligned and rearranged in sentences, for a total of about 3.5M tokens per language. Talks have been partitioned in train, development and test sets similarly in all respects to the MT tasks of the IWSLT 2016 evaluation campaign. In addition to describing the benchmark, we list the problems encountered in preparing it and the novel methods designed to solve them. Baseline MT results and some measures on sentence length are provided as an extrinsic evaluation of the quality of the benchmark.

Luisa Bentivogli, Arianna Bisazza, Mauro Cettolo et al.

Within the field of Statistical Machine Translation (SMT), the neural approach (NMT) has recently emerged as the first technology able to challenge the long-standing dominance of phrase-based approaches (PBMT). In particular, at the IWSLT 2015 evaluation campaign, NMT outperformed well established state-of-the-art PBMT systems on English-German, a language pair known to be particularly hard because of morphology and syntactic differences. To understand in what respects NMT provides better translation quality than PBMT, we perform a detailed analysis of neural versus phrase-based SMT outputs, leveraging high quality post-edits performed by professional translators on the IWSLT data. For the first time, our analysis provides useful insights on what linguistic phenomena are best modeled by neural models -- such as the reordering of verbs -- while pointing out other aspects that remain to be improved.