/Arabic-Image-Captioning

The concept of the project is to generate Arabic captions from the Arabic Flickr8K dataset, the tools that were used are the pre-trained CNN (MobileNet-V2) and the LSTM model, in addition to a set of steps using the NLP. The aim of the project is to create a solid ground and very initial steps in order to help children with learning difficulties.

Primary LanguageJupyter Notebook

Stargazers