← Concept IndexDEFINITION WHY IT MATTERS COMMONLY CONFUSED WITH
Video generation & temporal consistency
Also called: text-to-video, frame coherence
Generative video produces a sequence of frames that must stay consistent over time — the same object, lighting, and motion from one frame to the next, not just one good still.
Temporal consistency is the hard part and the tell: flicker, morphing objects, and physics that don't hold up are where generated video still breaks, which matters for both creation and detection.
A slideshow of independent images. The challenge is precisely the agreement between frames, which single-image generation never faces.