Addressing Artificial Intelligence Bias through Inclusivity: A Case Study with Nigerian Food Images

While the world remains fascinated by the technology and ingenuity behind Artificial Intelligence (AI), certain drawbacks prompt us to question its impact on humanity. One major drawback encountered is bias and lack of inclusivity. Recently, there have been instances where AI systems lacked equality, equity, and honesty in their results for specific operations. Given these events and the global advancement of AI applications, it is crucial to investigate the causes of bias in deployed models. To this end, we considered the model pretraining stage as the major cause of this challenge and identified the lack of non-Western datasets as a vital reason. This research aims to promote inclusivity by providing a food images dataset from Nigeria and a deep learning classification model trained on this dataset.