Meta’s researchers have developed a one-of-a-kind artificial intelligence-powered platform named Myosuite that can create realistic musculoskeletal models efficiently.
Meta collaborated with the University of Twente to develop Myosuite. The newly launched AI and ML-enabled platform can facilitate applications with real-life implications, such as rehabilitation, surgery, and shared autonomy assistive devices.
MyoSuite’s musculoskeletal models are up to 4,000 times faster than other simulators when it comes to addressing the data needs of modern machine learning algorithms.
Meta says that Apart from Myosuite, Meta is also releasing open-sourcing a set of comprehensive benchmarks for the ML community. This release will aid the community in continuing research in the field.
The newly launched platform can also be used to develop avatars for virtual worlds such as Horizon Worlds. “As we build for the metaverse, these models will help developers build avatars that move more realistically, making the experience more expressive and immersive,” mentioned Meta in a blog.
The advent of MyoSuite’s new technology, which intends to support modern-day features, will aid bioengineering and other related features in practice.
“We can train these models to perform complex movements like twirling a pen or rotating a key. We’re going to open source these models so researchers can use them to advance the field further,” said the CEO of Meta, Mark Zuckerberg.
The physiologically accurate musculoskeletal models developed by MyoSuite are orders of magnitude more computationally efficient and scalable than previous models. A set of well-tuned, physiologically correct musculoskeletal models that facilitate temporal interaction-rich musculoskeletal dynamics are included in the MyoSuite ecosystem.
The company hopes that its MyoSuite will become the pioneer platform for integrating motor and neural intelligence.