ED-T2V: An Efficient Training Framework for Diffusion-based Text-to-Video Generation | IEEE Conference Publication | IEEE Xplore