The filing points out that existing methods insert a generated frame between two real ones, which means waiting for both and adding delay. Sony’s version predicts forward from a single frame, with the goal of smoothing out hitching in heavy scenes without lowering visual quality. It can also predict several frames at once, and there’s a simpler non-AI fallback that shifts blocks of pixels along their motion vectors. Apparently, the model trains on the video stream itself, comparing each guess with the real next frame, so no human labeling is needed. However, the filing includes no latency or image quality numbers, and it remains pending, so there’s no guarantee any of it ships.
Frame generation would be a logical next step after the PS5 Pro’s PSSR upscaler, and leaks have already pointed to the PS6 targeting 4K at 120 FPS in most games. Multi-frame generation makes less sense on TVs that top out at 120 Hz, though, and some have raised concerns that developers could lean on it instead of optimizing their games. Sony has been on a roll with patents recently, with filings describing bosses that learn from your reaction times and retries, consoles that fix themselves using AI, an AI that fills in controller inputs for players with physical limitations, and even stream viewers controlling a cursor on the streamer’s screen. Like those, this one is probably best treated as something Sony is exploring for now.
