Inspiration
AA
What it does
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for Power Up
Built With
- nextjs
Power Up
A health and fitness app powered by chatgpt
Updates
Worked on secure OAUTH authentication via Google provided by Clerk Authentication Framework. I used Next.js along with Tailwind CSS for designing the frontend.
Leave feedback in the comments!
Log in or sign up for Devpost to join the conversation.
Pushed the frontend part to GitHub with integration completed. We enjoyed this hackathon as a whole and got exhausted as well lol.
Log in or sign up for Devpost to join the conversation.
Done with out last three features:
Conversational Refinement: Users can refine recommendations through conversation with LLaMA. For example, they can request recommendations for specific colours or styles.
Image Upscaling: If users are satisfied with the results, they can upscale the image from 512p to 2048p using ESRGAN_4x. This high-resolution image can then be sent to Google Lens for identifying and purchasing the generated clothing.
LoRA Models: We've also trained LoRA models for specific clothing items. This method can be used on a larger scale to train models based on an e-commerce website’s catalogue of clothing.
We are now starting to integrate the frontend with our backend
Log in or sign up for Devpost to join the conversation.
Done with our first two features:
Canvas Interaction: Users can interact with the canvas to add clothing items to images using the Stable Diffusion model. They can mask an area of the image and name the item of clothing they want to add.
Fashion Recommendations: Users can chat with our LLaMA model to receive fashion recommendations. LLaMA has access to the internet to search for relevant articles based on user queries and uses VGGNet for personalized recommendations based on past purchases.
Log in or sign up for Devpost to join the conversation.
Log in or sign up for Devpost to join the conversation.