2 papers
cs.SD2025
Semantic-Aware Confidence Calibration for Automated Audio Captioning
Lucas Dunker, Sai Akshay Menta, Snigdha Mohana Addepalli +1
Automated audio captioning models frequently produce overconfident predictions regardless of semantic accuracy, limiting their reliability in deployment. This deficiency stems from…
cs.SD2024
Source Separation & Automatic Transcription for Music
Bradford Derby, Lucas Dunker, Samarth Galchar +2
Source separation is the process of isolating individual sounds in an auditory mixture of multiple sounds [1], and has a variety of applications ranging from speech enhancement and…