1 paper
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…