dyelax/Adversarial_Video_Generation

A TensorFlow Implementation of "Deep Multi-Scale Video Prediction Beyond Mean Square Error" by Mathieu, Couprie & LeCun.

51
/ 100
Established

This project helps video game researchers and AI developers create clearer, more stable predictions of future video frames. Given a sequence of past video frames (like from Ms. Pac-Man gameplay), it generates the next frame in the sequence, producing sharper and more consistent visuals compared to traditional methods. Researchers studying generative models and video prediction will find this useful for evaluating adversarial techniques.

746 stars. No commits in the last 6 months.

Use this if you need to generate realistic, sharp future frames from a video sequence, especially for tasks where maintaining visual fidelity over time is crucial.

Not ideal if your primary goal is perfect, pixel-by-pixel accuracy with ground truth, as this focuses on visual plausibility and sharpness rather than exact matches.

video-prediction generative-models computer-vision-research video-game-analysis
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 25 / 25

How are scores calculated?

Stars

746

Forks

185

Language

Python

License

MIT

Last pushed

Oct 23, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/dyelax/Adversarial_Video_Generation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.