Traffic Sign Recognition, Edge Detection, Hand-written digit Recognition
- Category: Neural Networks, Computer Vision
- Client: CS50 Intro to AI, Harvard
- Timeline: Nov, 2020
CS50's AI course
Backpropagation and Gradient Descent, Computer Vision: Edge Detection, Convolutional Neural Networks (CNN), and Recurrent Neural Netowrks were discussed. I used a 3x3 kernel for edge detection on the photo of a bridge and a CNN (using Tensorflow library) to detect hand-written digits. Finally I designed a CNN to detect traffic signs from the gtsrb data. Watch the demo here.