1 paper
Benjamin Akera, Evelyn Nafula, Patrick Walukagga +3
The development of Automatic Speech Recognition (ASR) systems for low-resource African languages remains challenging due to limited transcribed speech data. While recent advances i…