- Visual Hippo is a python project where users can upload a picture of food and see restaurants near them that have that food, based off how similar the restaurants pictures are.
- Flask, Tensorflow, BS4, VGG6 Pretrained Model
- ScrapingBee Account for Key authentication
- Replicate AI accoubt for key authentication
- The project uses flask to connect back end to the front, and does not rlly implement a database in the MVC.
- Thus, the runtime of the tool is pretty slow. When a user uploads an image, a bunch of scraping scripts are called to scrape yelp for restaurants data and their images.
- Beautiful soup is used to parse the scraped html. However, due to IP blocking restrictions, the projects calls upon a scraper api to handle rotating proxies. That's the main bottleneck issue for the project.