MultiChoice and University of Pretoria produce world-class AI and machine learning technologies

MultiChoice South Africa partnered with the University of Pretoria to sponsor the MultiChoice Chair in Machine Learning, driving innovation and creating world-class scientists.

MultiChoice South Africa initially partnered with the University of Pretoria in 2018 to pursue collaboration and research.

MultiChoice recognised the key role artificial intelligence (AI), and specifically machine and deep learning, will play in unlocking a truly digital future.

Through its MultiChoice Chair in Machine Learning partnership, it invested in addressing the global skills shortage in machine learning and bridging the digital divide.

The Chair is jointly located in the University of Pretoria’s School of Information Technology and the School of Engineering.

The first five-year partnership was so successful that the agreement has been renewed for a second five-year term.

Prof Nelishia Pillay represents the Department of Computer Science, and Prof Pieter de Villiers represents the Department of Electrical, Electronic, and Computer Engineering.

The MultiChoice Chair in Machine Learning performs academic and applied research and human resource development in machine learning.

This research is applied to several advanced topics of interest to MultiChoice, making the most of the opportunities AI and machine learning offer.

MultiChoice has identified several opportunities and projects to apply AI and machine learning at points throughout its value chain.

This includes content creation, understanding what content to offer customers, customer service, and improving customer interactions.

These challenges and opportunities are addressed through master’s and PhD projects that propose novel and creative machine learning approaches to solve global problems.

It provides students access to real-world industry research and realistic data, helping them to solve real-world problems through AI.

Several exciting machine learning-related research projects are currently underway or have been completed through this partnership.

Topics include the automated genre labelling of motion picture trailers and computerised audio and video content analysis to extract meaningful and descriptive metadata.

They also created systems to automatically generate closed captions through automated audio processing and audio and video segmentation for scene skipping.

Students even produced automated movie trailer creation and forecasted weather-related impacts on DStv service delivery.

Some projects addressed big data challenges associated with the volume, velocity, variety, veracity, and value of video and audio content.

“Machine learning is a specialised field, and the partnership is a forward-looking way to nurture and improve these skills in South Africa,” Prof Pieter de Villiers said.

The benefits of this collaboration have also been realised by students who have performed vacation work at MultiChoice. Many were offered a permanent position at the company.