Translating speech with just images
Domain:
natural language processingRecord type:
paperdatasetmodel
Creator:
OneKam
Host:
Visually grounded speech models link speech to images. We extend this connection by linking images to text via an existing image captioning system, and as a result gain the ability to map speech audio directly to text. This approach can be used for speech translation with just images by having the audio in a different language from the generated captions. We investigate such a system on a real low-resource language, Yorùbá, and propose a Yorùbá-to-English speech translation model that leverages pretrained components in order to be able to learn in the low-resource regime. To limit overfitting, we find that it is essential to use a decoding scheme that produces diverse image captions for training. Results show that the predicted translations capture the main semantics of the spoken audio, albeit in a simpler and shorter form. Accepted at Interspeech 2024
Similar
SinFoS: A Parallel Dataset for Translating Sinhala Figures of SpeechMore than just warriors: Mythical and archetypal images of the hero in Swahili literatureTranslating into Morphologically Rich Languages with Synthetic PhrasesFine-Tuning Whisper for Speech Translation: A Case Study on Translating Darija to ArabicPepper(Mwendokasi) Images with AnnotationTranslating prose