Luma AI has launched Dream Machine which is another generative artificial intelligence (AI) that can generate images based on text input and images provided by users. Dream Machine can generate up to five seconds of video in various styles and camera movements.
Video generation using text is random and depends on the given prom. Generation using images is more controlled with the user uploading an image as a template and then giving instructions for the animation to be generated based on that image. Dream Machine can now be tried for free with it requiring login using Google account.
Based on our tests, videos take several minutes to generate at relatively low resolutions. The video result also still does not look so natural. Luma AI says Dream Machine is still not perfect as it still has problems understanding direction, natural way of moving, produces two-headed twins and cannot produce videos with text written in them.
