Machine learning is being used to ease the special effects workflow and generate new abstract visuals

A new music video featuring Chinese pop star Chris Lee,  ‘Rainy Day, But We Are Together,’ is the first of its kind to use artificial intelligence to generate dramatic special effects. After training machine learning algorithms to reconstruct a face in 3D and track its movement and expressions in real time, director Timothy Saccenti and a team from Intel were able to bypass the use of tracking markers and hyper-detailed video capturing guidelines.

While the special effects aren’t necessarily striving for hyperrealism, their means of production has various implications on how Intel and its AI experiments can progress special effects creation moving forward. In the future, the invisible ‘handy helper’ may become a new standard in videography, helping directors and cinematographers focus their creative energies on their artistic vision rather than the technical limitations of their projects.

Intel

A new music video featuring Chinese pop star Chris Lee,  ‘Rainy Day, But We Are Together,’ is the first of its kind to use artificial intelligence to generate dramatic special effects. After training machine learning algorithms to reconstruct a face in 3D and track its movement and expressions in real time, director Timothy Saccenti and a team from Intel were able to bypass the use of tracking markers and hyper-detailed video capturing guidelines.

While the special effects aren’t necessarily striving for hyperrealism, their means of production has various implications on how Intel and its AI experiments can progress special effects creation moving forward. In the future, the invisible ‘handy helper’ may become a new standard in videography, helping directors and cinematographers focus their creative energies on their artistic vision rather than the technical limitations of their projects.