Back to our Insights
Back to our Insights

Creating Diversity in Motion Capture: The Black Movement Project


Charlotte Mallo

Diversity and inclusion by design

Just a few months ago people from across the nation from all backgrounds took to the streets and social media in support of Black Lives Matter. In support of the movement, we wrote about how technology can empower diversity and racial equality through inclusive design

As a company, we have taken the time to look within, and also to those around us who are trailblazers in developing technology that is equally representative of the diversity in our world. How can we capture more diverse data sets? How does this affect technology such as hand tracking and gesture recognition?

To answer these questions we are highlighting the work of LaJuné McMillian, an incredibly talented new media artist and technologist, and her latest development, The Black Movement Project. 

“The Black Movement Project is not just about humanizing the data, it’s humanizing the technology. Are we reflecting the current state, and are we improving the reality?” said LaJuné. 

The Black Movement Project

The Black Movement Project is an online source of motion capture data from black performers and black character based models who are currently underrepresented in online databases. The mocap library LaJuné is creating will bring racial equality to virtual, augmented, and mixed reality projects, performances, games and research applications. 

Hand gestures and the way we perform them are deeply rooted in our history and culture, with subtle variations from one person to another offering a window into their world. 

This information speaks to our origins, culture, livelihood, race, gender and physical capabilities- and it’s only through access to this diverse information that it is possible to create hand tracking and gesture recognition software that is representative of and accessible to everyone in our society. 

Capturing Diverse Data 

LaJuné said it is not just a lack of access to tools and equipment for black people, but on a global scale, there’s a lack of awareness in how these tools operate. 

“With character building software and equipment it becomes very clear right off the bat who these tools are made for, even the size that the motion capture suits come in,” she said. 

Looking at online motion capture libraries, she observes a clear disconnect between a character and the movement that is available. “If I’m using motion capture data in my project for black people, I want to use data captured from black people,” said LaJuné. “It’s not just movement, it’s someone’s movement, and it’s important that we know their stories.”

In her project, LaJuné creates a profile on each individual who’s movement she captures, so that people know where and who the data comes from, and therefore, which data to use for their character models. 

Genesis, a live virtual reality performance based on motion capture

A Vision: Humanizing Data to Humanize Technology 

LaJuné’s vision for the Black Movement Project is to raise awareness of the need for education and diversity in data. She wants people to understand how data affects the technology we develop and therefore its accessibility. 

“For me, it’s about how to rehumanize the data, and it’s not just humanizing the data, it’s humanizing the technology. Are we reflecting the current state, and are we improving the reality?” said LaJuné. 

“In order to have diverse data, teams and technology, people need to have access to these tools even if they don’t have a gaming computer at home,” she said. She imagines a future library that connects anthropology and technology, educating communities and making the tools available so that everyone can understand and be a part of developing the technology that empowers our society. 

Machine Learning: Diverse Data at Clay AIR 

LaJuné’s Black Movement Project highlights the challenges we face in having access to diverse datasets to train our hand tracking and gesture recognition modules. 

With a critical eye towards the data we use, our R&D team generates a vast quantity and variety of hands (tattooed, accessories, shapes, colors, etc…) against numerous backgrounds, training the machine learning module that is responsible for hand tracking and gesture recognition. 

As a result, no matter what your hands look like, your gesture will be recognized in two frames or less, regardless of race or gender.  
We are always looking for diverse datasets to improve our technology. \

We’d also like to encourage you to support LaJuné ’s work via Patreon, where you can gain access to her patreon-only feed for blog posts and video updates, and be a part of creating a future that is free of racial discrimination.



Enabling next generation interactivity with digital interfaces through gesture controls.