Vectors in Speech Recognition

Feature vectors are used in speech recognition to identify phonetic sounds between each other. Usually, each sound clip is divided into several short snippets. Features are then extracted from each of these snippets, which are then used to identify between different sounds between the audio snippets.