CPT: Cross-Modal Prefix-Tuning for Speech-To-Text Translation | IEEE Conference Publication | IEEE Xplore