- Convolutional neural network based detection approach of undesirable SMS (Short Message Service) in the Cameroonian context.- Time aware implicit social influence estimation to enhance recommender systems performances.- Analysis of COVID-19 coughs: From the mildest to the most severe form, a realistic classification using deep learning.- SLCDeepETC : An On-Demand Analysis Ready Data pipeline on Sentinel-1 Single Look Complex for Deep Learning.- Robusness of image classification on imbalanced datasets using capsules networks.- Application of the multilingual acoustic representation model XLSR-53 for the transcription of Ewondo.- Self-supervised and multilingual learning applied to the Wolof, Swahili and Fongbe.- Explaining Meta-learner's Predictions: case of corporate CO2 emissions.

- Building and validating an ontology for public procurement : The case of Cameroon.- A Hybrid Algorithm Based on Tabu Search and K-means for Solving The Traveling Salesman Problem.- Hybridization Of A Recurrent Neural Network By Quadratic Programming For Combinatory Optimization.- A New Hybrid Algorithm Based on Ant Colony and Neural Networks with Attention Mechanism for Solving the Traveling Salesman Problem.- Recurrent Neural Network Parallelization for Hate Messages Detection.- Two high capacity text steganography schemes based on color coding.- Integration of the triple block data security model based on distributed crypto-steganography in a distributed system.