We are translating MS-COCO Captions to Luganda. This will solve two problems:
We get image captions in Luganda, useful to train a Luganda image-captioning AI. This is an artificial intelligence that is capable of describing an image you give it. We want to build a dataset that will enable students and researchers to build such AI models for Luganda, whenever they want.
We get sentences translated to Luganda. This will provide us with what AI researchers called 'aligned sentences'. This is very important for certain methods of English-Luganda translation, and is often useful for building an AI that can translate from English to Luganda.