NTU Researchers Unveil Upscale-A-Video: Pioneering Text-Guided Latent Diffusion for Enhanced Video Super-Resolution