ExpGest: Expressive Speaker Generation Using Diffusion Model and Hybrid Audio-Text Guidance | IEEE Conference Publication | IEEE Xplore