| Idea Name: |
Try n Buy E-Store |
| Slogan: |
Try it Before you Buy it |
| Supervisor Name: |
Dr. Ali Hassan |
| Supervisor Designation: |
Associate Professor |
| Supervisor School: |
CEME |
| Supervisor Department: |
Computer Engineering |
| Contact number: |
+923145222888 |
| Email ID: |
[email protected] |
| Abstract: |
| We are developing an e-commerce store for clothes. This store will have a try-on feature. the product (cloth item) the user will select, we will impose that cloth onto the user in real-time as well as any picture user will provide. Our app will predict the perfect size for the user (i.e M, L, etc). We will develop App as well as a website for this e-store. |
| What is the unmet need in society that your idea will fulfill ? |
| 56% of online users feel uncomfortable while buying clothes. 37% of users have doubts regarding the size. 20% of the clothes purchased end up being exchanged or returned. To lower these numbers and build more confidence among the buyer Try on feature is a perfect opportunity. |
| Who needs it ? How many would benefit ? |
| As the world is becoming a global village anyone can take advantage of this project. It will increase the trend of online shopping. It will eradicate traditional issues with online shopping. |
| How will the solution works |
| The problem will be divided into stages and will be solved step by step:
Dataset collection and Research. Develop and train models for "Pose Detection", "Human Body Parsing" and "U-net Segmentation". Develop a "Conditional GAN" model to use for "Transformation", "Stiching" and "Reinforcement". Pass outputs of 3 basic models with the input image to 3 variants of "Conditional GAN" model to get 3 final output images. Use image processing to combine these 3 output images into 1 desired final. |
| Who are your competitors ? How is your solution different |
| There is not a single e-commerce website that is using this technique. There are several ML models which are trying to implement this idea but none of them are attached to an e-commerce website. |
| Status: |
new |
| Entry Date & Time: |
2022-11-06 (0541) |
