Overview : This proposed system can help our agriculture firms to reduce production loss. This Project develops an automated approach to detect paddy diseases using a Deep Learning model based on the SMOTE-ENN resampling technique. We have used five popular Deep Learning algorithms: Convolutional Neural Network (CNN), VGG16, VGG19, Xception, and ResNet50. We measured the best model performance by their evaluation metrics. Comparing these five algorithms, Xception recorded the highest accuracy to detect paddy diseases. The model achieved an accuracy of 98%, precision of 97%, recall of 96%, and F1-Score of 97%.

Overview : Sign Language is a natural language used to communicate with deaf or dumb people. At present, communication pathways between well and disabled people are required, which demands resolving various specific issues, and one of those is gesture recognition. A technique for identifying static or dynamic gestures of the Bangla Sign Language is presented in this work. Mediapipe Holistic is used in this system as a feature extractor, and this sign recognition system is developed based on a deep learning approach named LSTM. The model had an accuracy of 98.51% and an F1 score value of 98.21% on the test data.