DejaVu: A Google Chrome Extension to Help Combat Visual Misinformation
Help journalists to search, check and annotate visual misinformation on the internet (deploying Deep Learning model in production)
![](social_tech_lab.png)
This project was built for the Social Technologies Lab at Cornell Tech
Academic Paper: DejaVu: A System for Journalists to Collaboratively Address Visual Misinformation
- Streamline the reverse image search process
- Index and search social media sources
- Support collaborative image annotation
![](architecture.png)
DejaVu System Architecture
Background Information:
According to First Draft, there exist seven types of mis- and disinformation on the internet:
- Satire of Parody: No intention to cause harm but has potential to fool
- False Connection: When headlines, visuals or captions don't support the content
- Misleading Content: Misleading use of information to frame an issue or individual
- False Context: When genuine content is shared with false contextual information
- Imposter Context: When genuine sources are impersonated
- Manipulated Content: When genuine information or imagery is manipulated to deceive
- Fabricated Content: New content is 100% false, designed to deceive and do harm
![](misinfo2.jpg)
Why Is This Type Of Content Being Created?