TA-V2A: Textually Assisted Video-to-Audio Generation | IEEE Conference Publication | IEEE Xplore