Playable Video Generation

TL;DR

This paper proposes a novel unsupervised framework for controllable, realistic video generation where users can select actions at each step, enabling interactive video creation without labeled data.

Contribution

It introduces a self-supervised, encoder-decoder model with a bottleneck action space for playable video generation, a new task in video synthesis.

Findings

Effective on diverse datasets with different environments

Achieves semantically consistent and realistic video outputs

Demonstrates controllability through user-selected actions

Abstract

This paper introduces the unsupervised learning problem of playable video generation (PVG). In PVG, we aim at allowing a user to control the generated video by selecting a discrete action at every time step as when playing a video game. The difficulty of the task lies both in learning semantically consistent actions and in generating realistic videos conditioned on the user input. We propose a novel framework for PVG that is trained in a self-supervised manner on a large dataset of unlabelled videos. We employ an encoder-decoder architecture where the predicted action labels act as bottleneck. The network is constrained to learn a rich action space using, as main driving loss, a reconstruction loss on the generated video. We demonstrate the effectiveness of the proposed approach on several datasets with wide environment variety. Further details, code and examples are available on our project page willi-menapace.github.io/playable-video-generation-website.