Revolutionizing Cataract Detection: A Machine Learning Approach
In this video, we take a deep dive into the world of machine learning and its potential to revolutionize cataract detection. Our team used a dataset of 700 images from Kaggle to train a model for cataract image classification. We increased the dataset by 30% using Gaussian Filter technique and visualized the data using various techniques like BGR, RGB, Gray Level, Adaptive Threshold Gray Level, and GLCM. We also used GLCM technique for feature extraction and extracted features like contrast, correlation, ASM, energy, dissimilarity, and homogeneity. We applied different models like KNN, Logistic Regression, Random Forest, Naive Bayes, and Support Vector Machine (SVM) and used techniques like Learning curve and Validation to select the best performing model. We also developed a web app using the Streamlit framework and deployed it on the Hugging Face website. The KNN model gave the best performance amongst all. With the help of machine learning, we can now improve the early detection and treatment of cataracts, saving countless lives from blindness.
Source Code : https://github.com/SajjadAli54/catara...
Live Website: https://huggingface.co/spaces/sajjada...