The need for training multilingual multi-task speech processing (MSP) models that perform both automatic speech recognition and speech-to-text translation is increasingly evident. However, a significant challenge arises from the conflicts among multiple objectives when using a single model. Multi-objective optimization can address this challenge by facilitating the optimization of multiple conflicting objectives and aligning the gradient updates in a common descent direction. While multi-objective optimization helps avoid conflicting gradient updates, a critical issue is that when there are many objectives, such as in MSP, it is often {\em difficult to find} a common descent direction. This leads to an important question: Is it more effective to separate highly conflicting objectives into different optimization levels or to keep them in a single level? To address this question, this paper investigates three multi-objective MSP formulations, which we refer to as \textbf{objective soup recipes}. These formulations apply multi-objective optimization at different optimization levels to mitigate potential conflicts among all objectives. To keep computation and memory overhead low, we incorporate a lightweight layer selection strategy that detects the most conflicting layers and uses only their gradients when computing the conflict avoidance direction. We conduct an extensive investigation using the CoVoST v2 dataset for combined multilingual ASR and ST tasks, along with the LibriSpeech and AISHELL-1 datasets for multilingual ASR, to identify highly conflicting objectives and determine the most effective training recipe among the three proposed multi-objective optimization algorithms.

How much audio is needed to fully observe a multilingual ASR model's learned sub-token inventory across languages, and does data disparity in multilingual pre-training affect how these tokens are utilized during inference? We address this question by analyzing Whisper's decoding behavior during inference across 49 languages. By logging decoding candidate sub-tokens and tracking their cumulative discovery over time, we study the utilization pattern of the model's sub-token space. Results show that the total number of discovered tokens remains largely independent of a language's pre-training hours, indicating that data disparity does not strongly influence lexical diversity in the model's hypothesis space. Sub-token discovery rates follow a consistent exponential saturation pattern across languages, suggesting a stable time window after which additional audio yields minimal new sub-token activation. We refer to this convergence threshold as acoustic saturation time (AST). Further analyses of rank-frequency distributions reveal Zipf-like patterns better modeled by a Zipf-Mandelbrot law, and mean sub-token length shows a positive correlation with resource level. Additionally, those metrics show more favorable patterns for languages in the Latin script than those in scripts such as Cyrillic, CJK, and Semitic. Together, our study suggests that sub-token utilization during multilingual ASR inference is constrained more by the statistical, typological, and orthographic structure of the speech than by training data scale, providing an empirical basis for more equitable corpus construction and cross-lingual evaluation.

ABSTRACT Multilingual automatic speech recognition (ASR) remains a challenging task, especially when balancing performance across high-and low-resource languages. Recent advances in sequence modeling suggest that architectures beyond Transformers may offer better scalability and efficiency. In this work, we introduce MLMA (Multilingual Language Modeling with Mamba for ASR), a new approach that leverages the Mamba architecture--an efficient state-space model optimized for long-context sequence processing--for multilingual ASR. Using Mamba, MLMA implicitly incorporates language-aware conditioning and shared representations to support robust recognition across diverse languages. Experiments on standard multilingual benchmarks show that MLMA achieves competitive performance compared to Transformer-based architectures.

Recent advancements in multilingual automatic speech recognition (ASR) have been driven by large-scale end-to-end models like Whisper. However, challenges such as language interference and expanding to unseen languages (language expansion) without degrading performance persist. This paper addresses these with three contributions: 1) Entire Soft Prompt Tuning (Entire SPT), which applies soft prompts to both the encoder and decoder, enhancing feature extraction and decoding; 2) Language-A ware Prompt Tuning (LAPT), which leverages cross-lingual similarities to encode shared and language-specific features using lightweight prompt matrices; 3) SPT - Whisper, a toolkit that integrates SPT into Whisper and enables efficient continual learning. Experiments across three languages from FLEURS demonstrate that Entire SPT and LAPT outperform Decoder SPT by 5.0% and 16.0% in language expansion tasks, respectively, providing an efficient solution for dynamic, multilingual ASR models with minimal computational overhead.

The deep integration of large language models and automatic speech recognition systems has become a promising research direction with high practical value. To address the overfitting issue commonly observed in Low-Rank Adaptation (LoRA) during the supervised fine-tuning (SFT) stage, this work proposes an innovative training paradigm Iterative LoRA Training (ILT) in combination with an Iterative Pseudo Labeling strategy, effectively enhancing the theoretical upper bound of model performance. Based on Whisper-large-v3 and Qwen2-Audio, we conduct systematic experiments using a three-stage training process: Focus Training, Feed Back Training, and Fix Training. Experimental results demonstrate the effectiveness of the proposed method. Furthermore, the MegaAIS research team applied this technique in the Interspeech 2025 Multilingual Conversational Speech Language Modeling Challenge (MLC-SLM), achieving 4th in Track 1 (Multilingual ASR Task) and 1st place in Track 2 (Speech Separation and Recognition Task), showcasing the practical feasibility and strong application potential of our approach.

This paper describes SHNU multilingual conversational speech recognition system (SHNU-mASR, team name-"maybe"), submitted to Track 1 of the INTERSPEECH 2025 MLC-SLM Challenge. Our system integrates a parallel-speech-encoder architecture with a large language model (LLM) to form a unified multilingual ASR framework. The parallel-speech-encoder consists of two pre-trained encoders, the Whisper-large-v3 encoder and mHuBERT -147 encoder. Their output embeddings are concatenated and fed into the LLM, enabling the model to leverage complementary acoustic and linguistic knowledge and achieve competitive performance. Moreover, we adopt a tri-stage training strategy to jointly update the low-rank adaptation modules and projector parameters of both the speech encoders and the LLM. In addition, we incorporate an additional language-aware prompt at the LLM input to enhance language-specific text generation. The SHNU-mASR system achieves an overall character/word error rate (CER/WER) of 11.76% on the blind evaluation set of the challenge, outperforming the official MLC-SLM baseline by 8.41 absolute CER/WER, without increasing the baseline training data.

Recent advancements in deep learning have significantly enhanced multilingual automatic speech recognition (ASR) due to the development of advanced model architectures and available large-scale multilingual datasets. Despite that, multilingual ASR still suffers from the curse of multilinguality in that different languages tend to interfere with each other, making it difficult for the ASR model to identify multiple languages effectively while sharing model capacity across them. This paper proposes an efficient finetuning framework for customized multilingual ASR via prepared LoRA language experts based on Whisper. Through LoRA expert fusion or knowledge distillation, our approach achieves better recognition performance on target languages than standard fine-tuning methods. Experimental results demonstrate that the proposed models yield approximately 10\% and 15\% relative performance gains in language-aware and language-agnostic scenarios, respectively.

We present a novel approach centered on the decoding stage of Automatic Speech Recognition (ASR) that enhances multilingual performance, especially for low-resource languages. It utilizes a cross-lingual embedding clustering method to construct a hierarchical Softmax (H-Softmax) decoder, which enables similar tokens across different languages to share similar decoder representations. It addresses the limitations of the previous Huffman-based H-Softmax method, which relied on shallow features in token similarity assessments. Through experiments on a downsampled dataset of 15 languages, we demonstrate the effectiveness of our approach in improving low-resource multilingual ASR accuracy.

While recent multilingual automatic speech recognition models claim to support thousands of languages, ASR for low-resource languages remains highly unreliable due to limited bimodal speech and text training data. Better multilingual spoken language understanding (SLU) can strengthen massively the robustness of multilingual ASR by levering language semantics to compensate for scarce training data, such as disambiguating utterances via context or exploiting semantic similarities across languages. Even more so, SLU is indispensable for inclusive speech technology in roughly half of all living languages that lack a formal writing system. However, the evaluation of multilingual SLU remains limited to shallower tasks such as intent classification or language identification. To address this, we present Fleurs-SLU, a multilingual SLU benchmark that encompasses topical speech classification in 102 languages and multiple-choice question answering through listening comprehension in 92 languages. We extensively evaluate both end-to-end speech classification models and cascaded systems that combine speech-to-text transcription with subsequent classification by large language models on Fleurs-SLU. Our results show that cascaded systems exhibit greater robustness in multilingual SLU tasks, though speech encoders can achieve competitive performance in topical speech classification when appropriately pre-trained. We further find a strong correlation between robust multilingual ASR, effective speech-to-text translation, and strong multilingual SLU, highlighting the mutual benefits between acoustic and semantic speech representations.

Approximately half of the world's population is multilingual, making multilingual ASR (MASR) essential. Deploying multiple monolingual models is challenging when the ground-truth language is unknown in advance. This motivates research efforts on configurable multilingual MASR models that can be prompted manually or adapted automatically to recognise specific languages. In this paper, we present the Configurable MASR model with Summary Vector (csvMASR), a novel architecture designed to enhance configurability. Our approach leverages adapters and introduces speech summary vector representations, inspired by conversational summary representations in speech diarization, to combine outputs from language-specific components at the utterance level. We also incorporate an auxiliary language classification loss to enhance configurability. Using data from 7 languages in the Multilingual Librispeech (MLS) dataset, csvMASR outperforms existing MASR models and reduces the word error rate (WER) from 10.33\% to 9.95\% when compared with the baseline. Additionally, csvMASR demonstrates superior performance in language classification and prompting tasks.