SEED Open Source - EA Official Site

SEED Open Source - EA Official Site SEED Research & Announcements Blogs Publications Open Source Careers Contact Us Research & Announcements Blogs Publications Open Source Careers Contact Us Open Source

SEED is committed to advancing game technology and contributing to the broader research community.

By providing open-source software, SEED enables researchers and developers to access cutting-edge tools and innovations. This initiative reflects SEED's dedication to giving back to the community, fostering collaboration, and driving progress in game development and interactive experiences.

Discover how SEED's contributions are shaping the future of gaming and technology.

AVA Capture

AVA Capture is a distributed system to control and record several cameras from a central UI. 

AVA Capture Constant-Time Stateless Shuffling and Grouping

Source code to accompany the related SEED blog post about using format-preserving encryption to shuffle items, or group them together in arbitrary group sizes.

Constant Time Stateless Dem Bones

An implementation of Smooth Skinning Decomposition with Rigid Bones, an automated algorithm to extract the linear blend skinning with bone transformations from a set of example meshes. 

Dem Bones Filter-Adapted Spatiotemporal Sampling for Real-Time Rendering

FastNoise generates noise textures optimized towards specific spatial and temporal filters, with specific per-pixel data types.

FastNoise GATA: Multi-Theme Generative Adversarial Terrain Amplification

Source code supporting the paper "Multi-Theme Generative Adversarial Terrain Amplification" from SIGGRAPH Asia 2019.

GATA GENEA Challenge

Data, code, and results from the GENEA Challenge, a benchmark for data-driven automatic co-speech gesture generation. Contributions to the code from SEED.

GENEA Gigi

A tool for rapidly prototyping rendering techniques as a node graph. Includes a visual editor, viewer, and compiler.

Gigi Machine Learning for Game Devs

Code repository for the related three-part SEED blog series that covers neural networks, weights & biases, and training learning systems.

ML for Game Devs Position-Based MPM

An accessible WebGPU implementation of Position-Based Material Point Method from SIGGRAPH 2024.

Material-Based MPM Project PICA PICA Assets

Assets used during the creation of Project PICA PICA, a real-time ray tracing experiment featuring self-learning agents.

Project PICA PICA Rig Inversion by Training a Differentiable Rig Function

An example of using the technique described in the paper "Rig Inversion by Training a Differentiable Rig Function" from SIGGRAPH Asia 2022.

Rig Inversion Towards Interactive Training of Non-Player Characters in Video Games

Two examples of Markov Ensemble as discussed in the paper "Towards Interactive Training of Non-Player Characters in Video Games" from the 2019 ICML Workshop on Human in the Loop Learning.

Training NPCs Game Library Subscribe EA app About Accessibility Help Legal & Privacy User Agreement Privacy & Cookie Policy Online Service Updates Security YouTube Terms of Service Google Privacy Policy

Optional Cookies, including those that support the Multimedia Player, are currently disabled based on Your Privacy Choices settings. 

Choose Accept and Play Video to enable the Multimedia Player, including its associated Cookies. This may involve the collection and sharing of your viewing information by YouTube for analytics and advertising purposes.

You can also choose to watch the video on YouTube.