| Idea Name: |
Memes Identifier |
| Slogan: |
AI Model to Identify Hateful-Memes |
| Supervisor Name: |
Dr. Usman Akram |
| Supervisor Designation: |
PhD Doctor |
| Supervisor School: |
CEME |
| Supervisor Department: |
Computer Engineering |
| Contact number: |
03336913921 |
| Email ID: |
[email protected] |
| Abstract: |
| A direct or indirect attack on people based on characteristics, including ethnicity, race, nationality, religion, caste, sex etc. And we are developing an AI Model to identify and remove this content |
| What is the unmet need in society that your idea will fulfill ? |
| Our idea has a great need in social media life, many people are spreading hate through social media apps and our idea will not show that content to normal people and this way we will be able to make peace among us. |
| Who needs it ? How many would benefit ? |
| As social media users are in billions so in this way it would be impactful to a very large community. |
| How will the solution works |
| We have data to build an AI Model, our data consist of images and extracted written on them. First we will do image captioning using different Deep Learning and Machine Learning models to give caption. And then we will combine both texts, caption and the text that was written on image. And then using another machine learning model to make a sentiment about the meme and our Google extention will blur that meme on social media in run time. |
| Who are your competitors ? How is your solution different |
| It's a Facebook launched competition, so our competitors are from all over the world, and from different places. |
| Status: |
new |
| Entry Date & Time: |
2021-01-11 (0345) |
