Francois Luus  Francois Luus photo         

contact information

Research Staff Member
IBM Research - Africa


Professional Associations

Professional Associations:  ACM  |  Golden Key International Honour Society  |  IEEE


Dr. Francois Luus is a Technical Lead in Machine Learning at IBM Research - Africa. He is responsible for research programmes in Advanced and Applied AI that solve difficult problems in the domains of Radio Astronomy and Computer Vision. He led a multi-year joint study with the Square Kilometre Array South Africa to develop machine learning for mitigation of radio frequency interference. Filed several joint patents and published on Generative ML solutions to RFI detection and removal. He was also a principal advisor to the SETI Institute in the areas of Cloud Computing and Cognitive Computing, where he has spearheaded analytics efforts for numerous observation campaigns. He has recently been working with a team developing Deep Learning solutions for Natural Language Understanding.

Notable projects at IBM Research include client/project management and sole development (in one month) of a fully-fledged interactive Twitter Sentiment Exploration web-platform capable of analyzing 100k tweets on a free Bluemix account, deployed during World Ecomobility Month for the City of Johannesburg, and for the University of Witwatersrand during the #FeesMustFall protests.

Consulting scientist and ML specialist for the SETI Institute since 2015, and previous advisor to the NASA Frontier Development Lab. Known for developing spectrogram folding 10x faster than existing SETI Institute solutions at the time. Contributed a powerful interactive supervision platform to TensorFlow to significantly reduce the data labelling expense - used for labeling SETI signals:

Advised and supported several Trappist-1 surveys with the SETI Institute, and performed various data processing on TB-scale measurements:

Previously, Francois was a consultant at the Remote Sensing Research Unit (CSIR, Meraka) where he developed robust domain adaptation for machine learning applied to land-use classification. He was also a researcher at the Sentech Chair in Broadband Wireless Multimedia Communications where he pioneered new information theoretic coding schemes for fast wireless networks.

Francois has done PhD work on applied machine learning in remote sensing, and he previously completed a B.Eng (Computer Engineering), B.Eng (hons) (Electronic Engineering) and an M.Eng (Electronic Engineering). Through his studies he has gained experience in artificial intelligence, machine learning, deep learning, computer vision, source/channel coding, wireless communications, wireless networking and remote sensing. Since 2008 he has presented on these subjects at numerous conferences, workshops and invited talks.