This paper presents the FBK contribution to the IWSLT-2024 ‘Test suites’ shared subtask, part of the Offline Speech Translation Task. Our contribution consists of the MuST-SHE-IWSLT24 benchmark evaluation, designed to assess gender bias in speech translation. By focusing on the en-de language pair, we rely on a newly created test suite to investigate systems’ ability to correctly translate feminine and masculine gender. Our results indicate that – under realistic conditions – current ST systems achieve reasonable and comparable performance in correctly translating both feminine and masculine forms when contextual gender information is available. For ambiguous references to the speaker, however, we attest a consistent preference towards masculine gender, thus calling for future endeavours on the topic. Towards this goal we make MuST-SHE-IWSLT24 freely available at: https://mt.fbk.eu/must-she/

This paper describes the FBK’s participation in the Simultaneous Translation Evaluation Campaign at IWSLT 2024. For this year’s submission in the speech-to-text translation (ST) sub-track, we propose SimulSeamless, which is realized by combining AlignAtt and SeamlessM4T in its medium configuration. The SeamlessM4T model is used ‘off-the-shelf’ and its simultaneous inference is enabled through the adoption of AlignAtt, a SimulST policy based on cross-attention that can be applied without any retraining or adaptation of the underlying model for the simultaneous task. We participated in all the Shared Task languages (English->German, Japanese, Chinese, and Czech->English), achieving acceptable or even better results compared to last year’s submissions. SimulSeamless, covering more than 143 source languages and 200 target languages, is released at: https://github.com/hlt-mt/FBK-fairseq/.

The paper describes the FBK submissions to the Subtitling track of the 2024 IWSLT Evaluation Campaign, which covers both the Automatic Subtitling and the Subtitle Compression task for two language pairs: English to German (en-de) and English to Spanish (en-es). For the Automatic Subtitling task, we submitted two systems: i) a direct model, trained in constrained conditions, that produces the SRT files from the audio without intermediate outputs (e.g., transcripts), and ii) a cascade solution that integrates only free-to-use components, either taken off-the-shelf or developed in-house. Results show that, on both language pairs, our direct model outperforms both cascade and direct systems trained in constrained conditions in last year’s edition of the campaign, while our cascade solution is competitive with the best 2023 runs. For the Subtitle Compression task, our primary submission involved prompting a Large Language Model (LLM) in zero-shot mode to shorten subtitles that exceed the reading speed limit of 21 characters per second. Our results highlight the challenges inherent in shrinking out-of-context sentence fragments that are automatically generated and potentially error-prone, underscoring the need for future studies to develop targeted solutions.

The attention mechanism, a cornerstone of state-of-the-art neural models, faces computational hurdles in processing long sequences due to its quadratic complexity. Consequently, research efforts in the last few years focused on finding more efficient alternatives. Among them, Hyena (Poli et al., 2023) stands out for achieving competitive results in both language modeling and image classification, while offering sub-quadratic memory and computational complexity. Building on these promising results, we propose ConfHyena, a Conformer whose encoder self-attentions are replaced with an adaptation of Hyena for speech processing, where the long input sequences cause high computational costs. Through experiments in automatic speech recognition (for English) and translation (from English into 8 target languages), we show that our best ConfHyena model significantly reduces the training time by 27%, at the cost of minimal quality degradation (∼1%), which, in most cases, is not statistically significant.

Despite its crucial role in research experiments, code correctness is often presumed solely based on the perceived quality of results. This assumption, however, comes with the risk of erroneous outcomes and, in turn, potentially misleading findings. To mitigate this risk, we posit that the current focus on reproducibility should go hand in hand with the emphasis on software quality. We support our arguments with a case study in which we identify and fix three bugs in widely used implementations of the state-of-the-art Conformer architecture. Through experiments on speech recognition and translation in various languages, we demonstrate that the presence of bugs does not prevent the achievement of good and reproducible results, which however can lead to incorrect conclusions that potentially misguide future research. As countermeasures, we release pangoliNN, a library dedicated to testing neural models, and propose a Code-quality Checklist, with the goal of promoting coding best practices and improving software quality within the NLP community.

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.

Streaming speech-to-text translation (StreamST) is the task of automatically translating speech while incrementally receiving an audio stream. Unlike simultaneous ST (SimulST), which deals with pre-segmented speech, StreamST faces the challenges of handling continuous and unbounded audio streams. This requires additional decisions about what to retain of the previous history, which is impractical to keep entirely due to latency and computational constraints. Despite the real-world demand for real-time ST, research on streaming translation remains limited, with existing works solely focusing on SimulST. To fill this gap, we introduce StreamAtt, the first StreamST policy, and propose StreamLAAL, the first StreamST latency metric designed to be comparable with existing metrics for SimulST. Extensive experiments across all 8 languages of MuST-C v1.0 show the effectiveness of StreamAtt compared to a naive streaming baseline and the related state-of-the-art SimulST policy, providing a first step in StreamST research.

The field of natural language processing (NLP) has recently witnessed a transformative shift with the emergence of foundation models, particularly Large Language Models (LLMs) that have revolutionized text-based NLP. This paradigm has extended to other modalities, including speech, where researchers are actively exploring the combination of Speech Foundation Models (SFMs) and LLMs into single, unified models capable of addressing multimodal tasks. Among such tasks, this paper focuses on speech-to-text translation (ST). By examining the published papers on the topic, we propose a unified view of the architectural solutions and training strategies presented so far, highlighting similarities and differences among them. Based on this examination, we not only organize the lessons learned but also show how diverse settings and evaluation approaches hinder the identification of the best-performing solution for each architectural building block and training choice. Lastly, we outline recommendations for future works on the topic aimed at better understanding the strengths and weaknesses of the SFM+LLM solutions for ST.

When this PhD started, the translation of speech into text in a different language was mainly tackled with a cascade of automatic speech recognition (ASR) and machine translation (MT) models, as the emerging direct speech translation (ST) models were not yet competitive. To close this gap, part of the PhD has been devoted to improving the quality of direct models, both in the simplified condition of test sets where the audio is split into well-formed sentences, and in the realistic condition in which the audio is automatically segmented. First, we investigated how to transfer knowledge from MT models trained on large corpora. Then, we defined encoder architectures that give different weights to the vectors in the input sequence, reflecting the variability of the amount of information over time in speech. Finally, we reduced the adverse effects caused by the suboptimal automatic audio segmentation in two ways: on one side, we created models robust to this condition; on the other, we enhanced the audio segmentation itself. The good results achieved in terms of overall translation quality allowed us to investigate specific behaviors of direct ST systems, which are crucial to satisfy real users’ needs. On one side, driven by the ethical goal of inclusive systems, we disclosed that established technical choices geared toward high general performance (statistical word segmentation of the target text, knowledge distillation from MT) cause an exacerbation of the gender representational disparities in the training data. Along this line of work, we proposed mitigation techniques that reduce the gender bias of ST models, and showed how gender-specific systems can be used to control the translation of gendered words related to the speakers, regardless of their vocal traits. On the other side, motivated by the practical needs of interpreters and translators, we evaluated the potential of direct ST systems in the “augmented translation” scenario, focusing on the translation and recognition of named entities (NEs). Along this line of work, we proposed solutions to cope with the major weakness of ST models (handling person names), and introduced direct models that jointly perform ST and NE recognition showing their superiority over a pipeline of dedicated tools for the two tasks. Overall, we believe that this thesis moves a step forward toward adopting direct ST systems in real applications, increasing the awareness of their strengths and weaknesses compared to the traditional cascade paradigm.

As part of the WMT-2023 “Test suites” shared task, in this paper we summarize the results of two test suites evaluations: MuST-SHEWMT23 and INES. By focusing on the en-de and de-en language pairs, we rely on these newly created test suites to investigate systems’ ability to translate feminine and masculine gender and produce gender-inclusive translations. Furthermore we discuss metrics associated with our test suites and validate them by means of human evaluations. Our results indicate that systems achieve reasonable and comparable performance in correctly translating both feminine and masculine gender forms for naturalistic gender phenomena. Instead, the generation of inclusive language forms in translation emerges as a challenging task for all the evaluated MT models, indicating room for future improvements and research on the topic. We make MuST-SHEWMT23 and INES freely available.

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 synchronized 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 speech translation 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.

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.

This paper describes the FBK’s participation in the Simultaneous Translation and Automatic Subtitling tracks of the IWSLT 2023 Evaluation Campaign. Our submission focused on the use of direct architectures to perform both tasks: for the simultaneous one, we leveraged the knowledge already acquired by offline-trained models and directly applied a policy to obtain the real-time inference; for the subtitling one, we adapted the direct ST model to produce well-formed subtitles and exploited the same architecture to produce timestamps needed for the subtitle synchronization with audiovisual content. Our English-German SimulST system shows a reduced computational-aware latency compared to the one achieved by the top-ranked systems in the 2021 and 2022 rounds of the task, with gains of up to 3.5 BLEU. Our automatic subtitling system outperforms the only-existing solution based on a direct system by 3.7 and 1.7 SubER in English-German and English-Spanish respectively.

Due to the complexity of bias and the opaque nature of current neural approaches, there is a rising interest in auditing language technologies. In this work, we contribute to such a line of inquiry by exploring the emergence of gender bias in Speech Translation (ST). As a new perspective, rather than focusing on the final systems only, we examine their evolution over the course of training. In this way, we are able to account for different variables related to the learning dynamics of gender translation, and investigate when and how gender divides emerge in ST. Accordingly, for three language pairs (en ? es, fr, it) we compare how ST systems behave for masculine and feminine translation at several levels of granularity. We find that masculine and feminine curves are dissimilar, with the feminine one being characterized by more erratic behaviour and late improvements over the course of training. Also, depending on the considered phenomena, their learning trends can be either antiphase or parallel. Overall, we show how such a progressive analysis can inform on the reliability and time-wise acquisition of gender, which is concealed by static evaluations and standard metrics.

In simultaneous speech translation (SimulST), finding the best trade-off between high output quality and low latency is a challenging task. To meet the latency constraints posed by different application scenarios, multiple dedicated SimulST models are usually trained and maintained, generating high computational costs. In this paper, also motivated by the increased sensitivity towards sustainable AI, we investigate whether a single model trained offline can serve both offline and simultaneous applications under different latency regimes without additional training or adaptation. Experiments on en->de, es show that, aside from facilitating the adoption of well-established offline architectures and training strategies without affecting latency, offline training achieves similar or better quality compared to the standard SimulST training protocol, also being competitive with the state-of-the-art system.

Gender bias is largely recognized as a problematic phenomenon affecting language technologies, with recent studies underscoring that it might surface differently across languages. However, most of current evaluation practices adopt a word-level focus on a narrow set of occupational nouns under synthetic conditions. Such protocols overlook key features of grammatical gender languages, which are characterized by morphosyntactic chains of gender agreement, marked on a variety of lexical items and parts-of-speech (POS). To overcome this limitation, we enrich the natural, gender-sensitive MuST-SHE corpus (Bentivogli et al., 2020) with two new linguistic annotation layers (POS and agreement chains), and explore to what extent different lexical categories and agreement phenomena are impacted by gender skews. Focusing on speech translation, we conduct a multifaceted evaluation on three language directions (English-French/Italian/Spanish), with models trained on varying amounts of data and different word segmentation techniques. By shedding light on model behaviours, gender bias, and its detection at several levels of granularity, our findings emphasize the value of dedicated analyses beyond aggregated overall results.

Simultaneous speech translation (SimulST) systems aim at generating their output with the lowest possible latency, which is normally computed in terms of Average Lagging (AL). In this paper we highlight that, despite its widespread adoption, AL provides underestimated scores for systems that generate longer predictions compared to the corresponding references. We also show that this problem has practical relevance, as recent SimulST systems have indeed a tendency to over-generate. As a solution, we propose LAAL (Length-Adaptive Average Lagging), a modified version of the metric that takes into account the over-generation phenomenon and allows for unbiased evaluation of both under-/over-generating systems.

Recent work has shown that systems for speech translation (ST) – similarly to automatic speech recognition (ASR) – poorly handle person names. This shortcoming does not only lead to errors that can seriously distort the meaning of the input, but also hinders the adoption of such systems in application scenarios (like computer-assisted interpreting) where the translation of named entities, like person names, is crucial. In this paper, we first analyse the outputs of ASR/ST systems to identify the reasons of failures in person name transcription/translation. Besides the frequency in the training data, we pinpoint the nationality of the referred person as a key factor. We then mitigate the problem by creating multilingual models, and further improve our ST systems by forcing them to jointly generate transcripts and translations, prioritising the former over the latter. Overall, our solutions result in a relative improvement in token-level person name accuracy by 47.8% on average for three language pairs (en->es,fr,it).

The primary goal of this FBK’s systems submission to the IWSLT 2022 offline and simultaneous speech translation tasks is to reduce model training costs without sacrificing translation quality. As such, we first question the need of ASR pre-training, showing that it is not essential to achieve competitive results. Second, we focus on data filtering, showing that a simple method that looks at the ratio between source and target characters yields a quality improvement of 1 BLEU. Third, we compare different methods to reduce the detrimental effect of the audio segmentation mismatch between training data manually segmented at sentence level and inference data that is automatically segmented. Towards the same goal of training cost reduction, we participate in the simultaneous task with the same model trained for offline ST. The effectiveness of our lightweight training strategy is shown by the high score obtained on the MuST-C en-de corpus (26.7 BLEU) and is confirmed in high-resource data conditions by a 1.6 BLEU improvement on the IWSLT2020 test set over last year’s winning system.

This project aimed at extending the test sets of the MuST-C speech translation (ST) corpus with new reference translations. The new references were collected from professional post-editors working on the output of different ST systems for three language pairs: English-German/Italian/Spanish. In this paper, we shortly describe how the data were collected and how they are distributed. As an evidence of their usefulness, we also summarise the findings of the first comparative evaluation of cascade and direct ST approaches, which was carried out relying on the collected data. The project was partially funded by the European Association for Machine Translation (EAMT) through its 2020 Sponsorship of Activities programme.

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 in direct 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%.

AbstractMachine translation (MT) technology has facilitated our daily tasks by providing accessible shortcuts for gathering, processing, and communicating information. However, it can suffer from biases that harm users and society at large. As a relatively new field of inquiry, studies of gender bias in MT still lack cohesion. This advocates for a unified framework to ease future research. To this end, we: i) critically review current conceptualizations of bias in light of theoretical insights from related disciplines, ii) summarize previous analyses aimed at assessing gender bias in MT, iii) discuss the mitigating strategies proposed so far, and iv) point toward potential directions for future work.

This paper describes FBK’s system submission to the IWSLT 2021 Offline Speech Translation task. We participated with a direct model, which is a Transformer-based architecture trained to translate English speech audio data into German texts. The training pipeline is characterized by knowledge distillation and a two-step fine-tuning procedure. Both knowledge distillation and the first fine-tuning step are carried out on manually segmented real and synthetic data, the latter being generated with an MT system trained on the available corpora. Differently, the second fine-tuning step is carried out on a random segmentation of the MuST-C v2 En-De dataset. Its main goal is to reduce the performance drops occurring when a speech translation model trained on manually segmented data (i.e. an ideal, sentence-like segmentation) is evaluated on automatically segmented audio (i.e. actual, more realistic testing conditions). For the same purpose, a custom hybrid segmentation procedure that accounts for both audio content (pauses) and for the length of the produced segments is applied to the test data before passing them to the system. At inference time, we compared this procedure with a baseline segmentation method based on Voice Activity Detection (VAD). Our results indicate the effectiveness of the proposed hybrid approach, shown by a reduction of the gap with manual segmentation from 8.3 to 1.4 BLEU points.

Speech translation (ST) has lately received growing interest for the generation of subtitles without the need for an intermediate source language transcription and timing (i.e. captions). However, the joint generation of source captions and target subtitles does not only bring potential output quality advantages when the two decoding processes inform each other, but it is also often required in multilingual scenarios. In this work, we focus on ST models which generate consistent captions-subtitles in terms of structure and lexical content. We further introduce new metrics for evaluating subtitling consistency. Our findings show that joint decoding leads to increased performance and consistency between the generated captions and subtitles while still allowing for sufficient flexibility to produce subtitles conforming to language-specific needs and norms.

Transformer-based models have gained increasing popularity achieving state-of-the-art performance in many research fields including speech translation. However, Transformer’s quadratic complexity with respect to the input sequence length prevents its adoption as is with audio signals, which are typically represented by long sequences. Current solutions resort to an initial sub-optimal compression based on a fixed sampling of raw audio features. Therefore, potentially useful linguistic information is not accessible to higher-level layers in the architecture. To solve this issue, we propose Speechformer, an architecture that, thanks to reduced memory usage in the attention layers, avoids the initial lossy compression and aggregates information only at a higher level according to more informed linguistic criteria. Experiments on three language pairs (en→de/es/nl) show the efficacy of our solution, with gains of up to 0.8 BLEU on the standard MuST-C corpus and of up to 4.0 BLEU in a low resource scenario.

Automatic translation systems are known to struggle with rare words. Among these, named entities (NEs) and domain-specific terms are crucial, since errors in their translation can lead to severe meaning distortions. Despite their importance, previous speech translation (ST) studies have neglected them, also due to the dearth of publicly available resources tailored to their specific evaluation. To fill this gap, we i) present the first systematic analysis of the behavior of state-of-the-art ST systems in translating NEs and terminology, and ii) release NEuRoparl-ST, a novel benchmark built from European Parliament speeches annotated with NEs and terminology. Our experiments on the three language directions covered by our benchmark (en→es/fr/it) show that ST systems correctly translate 75–80% of terms and 65–70% of NEs, with very low performance (37–40%) on person names.

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.

In automatic speech translation (ST), traditional cascade approaches involving separate transcription and translation steps are giving ground to increasingly competitive and more robust direct solutions. In particular, by translating speech audio data without intermediate transcription, direct ST models are able to leverage and preserve essential information present in the input (e.g.speaker’s vocal characteristics) that is otherwise lost in the cascade framework. Although such ability proved to be useful for gender translation, direct ST is nonetheless affected by gender bias just like its cascade counterpart, as well as machine translation and numerous other natural language processing applications. Moreover, direct ST systems that exclusively rely on vocal biometric features as a gender cue can be unsuitable or even potentially problematic for certain users. Going beyond speech signals, in this paper we compare different approaches to inform direct ST models about the speaker’s gender and test their ability to handle gender translation from English into Italian and French. To this aim, we manually annotated large datasets with speak-ers’ gender information and used them for experiments reflecting different possible real-world scenarios. Our results show that gender-aware direct ST solutions can significantly outperform strong – but gender-unaware – direct ST models. In particular, the translation of gender-marked words can increase up to 30 points in accuracy while preserving overall translation quality.

This paper describes FBK’s participation in the IWSLT 2020 offline speech translation (ST) task. The task evaluates systems’ ability to translate English TED talks audio into German texts. The test talks are provided in two versions: one contains the data already segmented with automatic tools and the other is the raw data without any segmentation. Participants can decide whether to work on custom segmentation or not. We used the provided segmentation. Our system is an end-to-end model based on an adaptation of the Transformer for speech data. Its training process is the main focus of this paper and it is based on: i) transfer learning (ASR pretraining and knowledge distillation), ii) data augmentation (SpecAugment, time stretch and synthetic data), iii)combining synthetic and real data marked as different domains, and iv) multi-task learning using the CTC loss. Finally, after the training with word-level knowledge distillation is complete, our ST models are fine-tuned using label smoothed cross entropy. Our best model scored 29 BLEU on the MuST-CEn-De test set, which is an excellent result compared to recent papers, and 23.7 BLEU on the same data segmented with VAD, showing the need for researching solutions addressing this specific data condition.