Undergraduate Thesis: Accomplished the One-Shot Learning technique by developing a Siamese Neural Network and Triplet Loss function for speech recognition tasks that could classify a new untrained speaker after receiving their audio data just once. It outperformed the traditional methods and achieved impressive accuracy on a limited dataset.
P. Nowshin, S. R. Dipto, I. Ahmed, and et al., “One voice is all you need: A one-shot approach to recognize your voice,” in 2022 7th International Conference on Data Science and Machine Learning Applications (CDMA), pp. 103–108, IEEE, 2022