👤 Ian Soboroff Person

Philip Resnik argues that evaluating AI speech-to-text systems solely by overall accuracy or word error rate overlooks the potentially severe consequences of particular mistakes. He cites a study in w