Image recognition is a widely used technique to enable computers to recognise the objects inside a graphical input. It is used by a large amount of tasks including image labelling, image search, medical disease recognition, and self-driving cars. For a long time, the best algorithm to use was a Support Vector Machine with several optimisations. This changed in 2012 when an annual ImageNet Large Scale Visual Recognition Challenge was won by a convolutional neural network. Since then a large progress has been seen in the area. The objective of this project was to analyse several image recognition techniques with the main focus on the effectiveness of ResNet architecture.
This project is accompanied with a report that can be found here.
The confusion matrix depicting the model accuracy
