Promptable Game Models: Text-Guided Game Simulation via Masked Diffusion Models

ACM Transactions on Graphics

Willi Menapace☨, Aliaksandr Siarohin, Stéphane Lathuilière, Panos Achlioptas,
Vladislav Golyanik, Sergey Tulyakov, Elisa Ricci

☨ Work performed while interning at Snap Inc.

Overview Dataset
Action-Conditioned Video Prediction Video Completion Unconditional Generation Random Conditioning Baselines Ablation
Camera Manipulation Style Manipulation Baselines Ablation
Paper GitHub

Ablation results for the synthesis model

We evaluate our synthesis model under different configuration to validate our design choices.

Ours w/o enhancer

Note the lack of shadows below players and imprecise colors of objects, eg. the ball.

Ours w/o human deformation

Note artifacts in the hands of the player at the top.

Ours w/o planes

Note the lack of detail in the judge stand and top left and top right corners of the image.

Ours w/o voxels

Note the artifacts in the geometry of the player, eg. duplicated right leg

Ours w/o multiobject encoder

Note the lack of details in the scene with respect to our method, eg. in the judge stand

Ours small

The full version of our model, trained with a reduced amount of computational resources, matching the one used for the ablation experiments.

Ours

The full version of our model