Nikhil Maurya

Identifying Similar Images in the Item Bank: Build a CNN Model to get similar X-Ray images. Used Auto Encoder-Decoder for feature extraction of images. Clustered images using K-Means. Used KNN for identifying similar images. Speech Emotion Recognition for Quality Check of Proctors: Data Preparation. Extracted features of audio using MFCC. Done data augmentation. Build CNN-LSTM model for predicting emotion of proctors. Object Detection-Enhancement in the existing model for the remote proctoring: Collected images using google image scraper. Done data annotation. Added new classes to existing Object Detection Model.

Classification of disease in a rice leaf: Detected the different types of diseases of Rice leaf. Used Pretrained CNN model. Done data augmentation and image preprocessing. The validation accuracy of the model is 80%. Insurance Claim Prediction: Predicted that a driver will claim insurance or not. Developed this POC project as a part of internship. Used Logistic Regression, Random Forest, XGBoost classifier and SVM. Done Hyperparameters Tuning using Grid Search CV. The best accuracy is 98% with XGBoost.

Conducting and delivery of mathematics sessions for class 10th to 12th students. Handhold students who struggle to understand the complex concept of Mathematics.