Abstract
A system and method are disclosed that automatically trigger system-generated captions for images. The system includes a machine learning model and an interface that proposes captions for an image that a user intends to share. The machine learning model is generated by considering the factors such as: extraction of metadata from the image, semantic information extracted from raw image, context of the current chat, image captions from previously shared images, personal context or user modifications to the image. Based on the results of the machine learning model, the system generates and proposes several captions to the user. The user then selects a suitable caption and shares the image. The user may also modify or manually enter the caption. The system generated image captioning enables easy sharing of captioned images which are personalized and artistic.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Feuz, Sandro and Millius, Sebastian, "Automatic Context-Aware Image Captioning", Technical Disclosure Commons, (May 23, 2017)
https://www.tdcommons.org/dpubs_series/532