Projects

Handwriting Detection System using Machine Learning

Sep 2023 - Nov 2023

• Enhanced Image Analysis with OpenCV and Pytesseract Integrated OpenCV for comprehensive image preprocessing, including resizing, noise reduction, and contrast enhancement to improve the clarity and consistency of input images. Leveraged Pytesseract for Optical Character Recognition (OCR), enhancing the accuracy of image text analysis by 30%. This combination helped streamline the model's ability to accurately interpret and extract text from images across different conditions, such as varying lighting and resolution levels.
• Developed a robust machine learning model capable of recognizing and interpreting handwritten text across diverse document types with an accuracy of 85%. The model was trained on a rich dataset of handwritten samples, covering various styles and formats to ensure adaptability and precision. Techniques such as data augmentation and hyperparameter tuning were applied to improve model generalization and accuracy on both familiar and new data samples.
• Implemented a user-friendly, interactive web application using Streamlit to provide a real-time interface for users. The application allows users to upload documents and immediately view OCR results with a response time of under 2 seconds. Optimized the processing pipeline to handle concurrent requests efficiently, ensuring that the system is both scalable and responsive for end-users, with an intuitive design that requires no technical expertise to use.
• Conducted an extensive literature review of over 17 research papers in the fields of OCR, image processing, and handwriting recognition. Synthesized the findings into a detailed project report covering the architecture, model performance, and comparative analysis with other solutions. Additionally, the report outlined future improvements, including potential integration of more advanced neural network architectures, and discussed challenges and proposed solutions for increasing model robustness and accuracy.
• Get the project: click here

Guess The Number Game

Nov 2023 - Dec 2023

• In this game a number is generated randomly, You have to guess that number.
• You have 10 attempts to guess.
• At every attempts you have given hints like your number is more or less than the randomly generated number.
• If you failed all the attempts, click on "Start New Game" to start a new game.
• Get the Game: click here

Facial Emotion Recognition Using Machine Learning

Mar 2023 - Jun 2023

• This involves building a visually engaging and interactive user interface (UI) using modern web development tools—HTML5 for structuring content, CSS3 for styling, and JavaScript for dynamic, responsive interactions. Making the design responsive means ensuring the layout and elements adapt seamlessly to various.
• Processed and normalized a dataset of over 40,000 images for emotion recognition, applying data augmentation to improve model robustness.
• Trained a machine learning model on a Kaggle dataset, reaching 80% accuracy in classifying six core emotions—fear, anger, disgust, happiness, neutrality, and sadness—using feature extraction and optimization techniques.
• This entails training a machine learning model specifically designed for emotion classification, using a curated dataset from Kaggle. The model’s task is to classify images into six emotions: fear, anger, disgust, happiness, neutrality, and sadness. Achieving 80% accuracy reflects a relatively strong performance in this domain, which often involves complex visual patterns. Feature extraction methods likely included techniques such as convolutional filters to highlight critical features in the images (like facial expressions) that are relevant for emotion detection.
• Get the Project: click here