Inspiration

Our team observed that most small businesses are struggling with their digital marketing strategy since they don’t usually have the know-how needed for it. In addition, they count a lot in this digital age on socials medias to generate traffic and bring in customers. However socials media works on the principle of maximizing interactions. This usually results in the case of small businesses in less views and traffic since if their posts are not generating interactions then they will recommended less. That’s why our goal is to transform the digital marketing for small business and making the process of creating creatives, interesting post that generate more interactions as smooth as possible. For this we developed a model that helps business owners generate creative pictures starting from a source picture such as a picture of a product. Our project generates various images depending on the users input text personalized to his product.

How we built it

We created a novel pipeline which combines automatic image segmentation based on a text prompt with state of the art image generation.

Challenges we ran into

Lack of GPUs for fast inference and higher resolution images.

Accomplishments that we're proud of

Building and deploying the inference pipeline successfully.

Built With

Share this project:

Updates