Keywords: Image classification, Interactive User Interface, Face detection, Image processing, Preprocessing, Visualization, Optimization Github

1. Created a user-friendly web service with an optimized image classifier that accurately categorizes images of 5 business leaders, showcasing their classification scores on the User Interface

2. Collected raw images of 5 business leaders, namely, Anand Mahindra, Elon Musk, Gautam Adani, Jeff Bezos, and Ratan Tata using the "Fatkun Batch Download Image" Chrome extension

3. Utilized haarcascade algorithm to detect faces and eyes in the images, then cropped out the detected faces to generate a dataset for classification

4. Performed feature extraction with wavelet transform and labeled the images of business moguls with numbers 0 to 4 to identify patterns and make accurate predictions

5. The Support Vector Classifier, optimized with gridSearchCV, achieved an adequate score of approximately 80%, outperforming the Random Forest model

Fig. 1. Face detection with haarcascade algorithm

Fig. 2. Feature extraction with Wavelet transform

Fig. 3. The user interface correctly identifies images of leaders, displaying both the name and the probability score