Welcome to the Celeb Recognition Web App! With our application, you can upload images or even use a webcam to recognize celebrities. Built using state-of-the-art techniques, we ensure efficient and fast recognition.
Notice the available celebs for recognition - full list
- Make sure you have python installed and install the relevant imported packages from the code if needed.
- Run
serverfile to start the server. - Go to http://127.0.0.1:8000/.
- If you encounter any troubles feel free to contact me.
Check out the demo!
-
The tested images were took by an iphone, capturing a computer screen:
-
Then, I captured an image using the webcam.
-
DataBase: All the celebrity data was meticulously scraped from IMDb.
-
AWS RDS Database Integration: Our app uses AWS RDS to store data across multiple tables:
- Celeb2: Contains detailed information about celebrities.
- Celeb_Images: Houses URLs for additional celebrity images.
- Face_Encodings: Contains face encodings for all processed images.
-
AWS S3 Storage: Main images of celebrities are stored in AWS S3. We've also implemented a mechanism to generate special URLs to present these images directly from the database.
-
Performance: For optimal performance in image similarity search, we employ the Annoy library.
-
Multiple Images: Each celebrity is backed by 10 - 30 images in our database to ensure accurate recognition.
-
Webcam Support: Users can utilize their webcams for real-time celebrity recognition. However, due to the downgraded resolution of webcam images, the accuracy might be slightly lower than using uploaded images.
-
Analytics: Using Google Analytics for detailed statistics and analytics about website traffic and conversions.
-
AWS EC2: Integration with EC2 will be available in the upcoming version. Stay tuned for the website url!
-
More Celebrities: Our database is ever-growing! Stay tuned for the addition of more celebrities in the next update.