Effectively combining Phi-4 and NLLB for Spoken Language Translation: SPRING Lab IITM’s submission to Low Resource Multilingual Indic Track

This paper presents the methodologies implemented for Spoken Language Translation for the language pairs Hindi-English, Bengali-English and Tamil-English for the Low Resource Multilingual Indic Track of The International Conference on Spoken Language Translation (IWSLT) for 2025.We adopt a cascaded approach and use a fine-tuned Phi-4 multimodal instruct model for Automatic Speech Recognition(ASR) and a fine-tuned NLLB model for Machine Translation(MT).Finally, we discuss targeted solutions (e.g.data augmentation, multilingual training, targeted finetuning) to boost low-resource translation, noting that significant retraining on additional Tamil data is likely needed.