MIT Researchers Predict The Future From Still Photos
Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), possibly under the influence of science fiction novelists, have been working on generating pictures of the future from ordinary still photographs. I say possibly, because surely no one would write about a camera that took a still picture at one moment in time, that could be developed into a video of the future. Nutty idea, don't you think?
(MIT creating videos of the future from still images)
We capitalize on large amounts of unlabeled video in order to learn a model of scene dynamics for both video recognition tasks (e.g. action classification) and video generation tasks (e.g. future prediction). We propose a generative adversarial network for video with a spatio-temporal convolutional architecture that untangles the scene's foreground from the background. Experiments suggest this model can generate tiny videos up to a second at full frame rate better than simple baselines, and we show its utility at predicting plausible futures of static images. Moreover, experiments and visualizations show the model internally learns useful features for recognizing actions with minimal supervision, suggesting scene dynamics are a promising signal for representation learning. We believe generative video models can impact many applications in video understanding and simulation.
A bright light seared in to being, vanished; Chuck, blinded, peered and then saw, standing in the center of the room with a camera in his hands, a man he recognized. Recognized and disliked.
"Hello, Chuck," Bob Alfson said... "This film I'm using - I'm sure you've run across it at CIA; it's expensive, but helpful." He explained to both Chuck and Joan. "I've just taken an Agfom potent-shot. Does that strike a chord? What I have in this camera is not a record of what you did just now but what will go on here in the next half hour..."
(Read more about the Agfom potent shot film)