Automatic image captioning model based on Caffe, using features from bottom-up attention.
Primary LanguageJupyter NotebookMIT LicenseMIT