Speaker Recognition System
Speaker recognition systems aim to automatically identify individuals based on their voice, with applications ranging from security systems to virtual assistants. Current research heavily focuses on improving robustness against adversarial attacks (e.g., voice morphing, spoofing) and addressing biases stemming from training data limitations, often employing deep neural networks like ResNets, ECAPA-TDNN, and transformer-based models such as WavLM. These advancements are crucial for enhancing the reliability and security of voice-based biometric systems, impacting fields like forensics, healthcare, and personal security.
Papers
May 27, 2022
April 27, 2022
April 25, 2022
April 8, 2022
April 1, 2022
March 28, 2022
March 16, 2022
March 10, 2022
February 23, 2022
February 2, 2022
January 9, 2022
November 3, 2021