Project Report
Browse
43 results
Search Results
Item An Unified Quantum Classical Model For Noisy Label Medical Image Binary Classification.(Comilla University, 1-Feb-2025) Bhuiyan, Taki Jakera; Fahim , Jahid KarimThe presence of noisy labels in medical imaging datasets can severely impact diagnosticItem TRAFFIC SIGN DETECTION UNDER ADVERSE CONDITIONS(Comilla University, 1-Feb-2025) Mohiuddin, Mohammad; Hasan, MehediAutomatic traffic sign detection and recognition play a crucial role in real-world applications such as autonomous driving systems and driver assistance technologies. This project aims to develop a system using deep learning techniques to enhance road safety and assist autonomous vehicles by automatically detecting road signs and signals in adverse situations such as intense lighting, snowfall, and rainy conditions. In this project, we utilize a Convolutional Neural Network (CNN) algorithm to train the model using a large dataset provided by the German Traffic Sign Recognition Benchmark (GTSRB). Additionally, OpenCV is employed for image processing, while TensorFlow/Keras is used for model training and image classification. The proposed model achieves an accuracy of 95.33% on the test dataset, demonstrating its robustness in detecting traffic signs under challenging conditions such as intense lighting, foggy weather, or heavy snowfall. This system can be integrated into driver assistance systems and autonomous vehicle. In the future, it can be further improved for real-time traffic sign detection and enhanced to achieve higher accuracy in complex environmentItem A Time-Nonlocal Optimization Approach for Image classification(Comilla University, 1-Feb-2025) Das, Hridoy Chandra; Hafsa, AkterImage classification remains a fundamental problem in artificial intelligence, with applicationsItem SYMPTOM-BASED DISEASE PREDICTION AND PRECAUTION SYSTEM USING MACHINE LEARNING & DEEP LEARNING(Comilla University, 1-Feb-2025) Hasan , Md FuadThe joining of machine learning and common dialect handling (NLP) has changed healthcare, empowering mechanized frameworks for foreseeing and diagnosing maladies. This consideration pointed to create an progressed system for malady classification utilizing literary portrayals of indications. The system's establishment may be a K-nearest neighbor (KNN) classifier, chosen for its direct approach and demonstrated viability in classification errands. Side effect portrayals are preprocessed utilizing TF-IDF vectorization, changing over content into numerical information reasonable for machine learning calculations. To progress information comprehension, exploratory information examination (EDA) strategies, counting word cloud visualization and stop word disposal, are utilized. The inquiry examines modern Neural Arrange structures, such as CNN, Bi-LSTM, GRU, LSTM, CNN+GRU, RNN, and CNN+LSTM, nearby routine machine learning models like Back Vector Machine (SVM), Calculated Relapse (LR), Choice Tree (DT), Multinomial Gullible Bayes (MNB), and Irregular Timberland Classifier (RFC). Besides, transformer-based models, counting BERT and DISTILBERT, are utilized to tackle state-of-the-art NLP capabilities. Execution evaluation was conducted utilizing exactness measurements, classification reports, and perplexity frameworks, with the Choice Tree (DT) demonstrate developing as the best entertainer. To improve availability, a Gradio-powered intuitively chatbot was made, empowering clients to input symptoms and get infection forecasts in conjunction with preventive measures. This inquires about underscores the potential of coordination NLP methods with machine learning to upgrade symptomatic accuracy and productivity in healthcare. By combining prescient analytics with user-friendly interfacing, this extend contributes to the extending field of AI-driven restorative arrangements, advertising a versatile and viable device for infection determination and quiet care. Social media stages have advanced into a essential source of news utilization for groups of onlookers. The expansion of deluding data over conventional media sources, counting social organizing locales, news web journal posts, and online gatherings, has made it challenging to recognize valid news sourcesItem Social Media-Based Psychological Insights for Detecting Depression Using Comparative Analysis of Machine Learning Algorithm(Comilla University, 1-Feb-2025) Nowar, SohanaPeople's communication methods are changing due to rapid advancements in technology. As the Internet has grown, social media platforms like Facebook, Instagram, Telegram, and Twitter have gained prominence as places for people to express their feelings, ideas, and psychological patterns. Millions of people worldwide suffer from depression, a common and devastating mental illness. Effective intervention and therapy depend on early discovery and precise diagnosis. With an emphasis on both conventional methods like clinical interviews and standardized tests as well as more recent approaches that use machine learning and natural language processing (NLP), this study examines a variety of psychological analysis tools for diagnosing depression. Through text analysis, psychological analysis gleans facts, features, and significant information from user perspectives. Social networks are used by psychological analytic researchers to identify activity and behavior associated with sadness. Social networks convey a wealth of information on the attitudes and behaviors associated with the onset of depression, including low sociability, medical treatment, self-centeredness, and a high level of daytime and nighttime activity. The study emphasizes how behavioral indicators, speech patterns, facial expressions, and textual analysis from online interactions or patient interviews can be used to detect signs of depression. The project intends to provide a thorough framework for identifying depression by evaluating psychological data using both qualitative and quantitative methods. This could increase diagnostic precision and result in more individualized treatment regimens. The effectiveness and accessibility of mental health care could be greatly improved by incorporating these strategies into clinical practice. In order to detect depression in tweets, we employed five machine learning classifiers in this research: logistic regression, K-nearest neighbor, support vector machines, decision trees, and LSTM. Technically, the oversampling of approaches is examined in two types of datasets: balanced and imbalanced. The findings indicate that for both balanced and imbalanced information, the LSTM classification model performs better than the other baseline models in the healthcare approach to depression being recognized.Item Smart Vehicle Number Plate Detection System(Comilla University, 1-Feb-2025) Latif, Md. AbdulThe implementation of number plate recognition systems has significantly improved the city's traffic conditions. These systems provide guidance for creating an efficient intelligent transportation system. With the rapid growth in vehicle numbers, Automatic Number Plate Recognition (ANPR) has become a crucial tool for traffic management. ANPR plays a vital role in traffic and security surveillance, utilizing advanced technology and image processing techniques to automatically identify characters on vehicle license plates.Item ROAD LANE LINE DETECTION USING U-NET ARCHITECTURE FOR SELF DRIVING CARS(Comilla University, 2-Feb-2025) Akter, Mosammat Sania; Mouri, Ifrat JahanThis project introduces a deep learning-based road lane line detection system using the U-Net architecture, aimed at enhancing autonomous driving and advanced driver assistance systems (ADAS). The model is trained on the CULane dataset, leveraging semantic segmentation to accurately detect lane markings in diverse road conditions. Preprocessing techniques such as resizing, normalization, and data augmentation are applied to improve model robustness. The system employs binary cross-entropy loss and is optimized using the Adam optimizer for efficient learning. A custom data pipeline ensures smooth training and evaluation, with real-time predictions visualized for performance assessment. The trained model effectively identifies lane boundaries, even under challenging lighting and occlusion scenarios. Extensive experiments validate the system’s accuracy and reliability in real-world conditions. The model achieves an accuracy of 96.15%, outperforming many existing lane detections models. The proposed approach contributes to safer and more intelligent vehicle navigation by reducing lane departure risks. Future enhancements include integrating temporal consistency for improved detection stability.Item "RNN's LSTM Based Deep Learning Model for Stock Price Prediction"(Comilla University, 1-Feb-2025) Mozumder, Minhajul IslamOne of the most significant activities in the world of finance is stock trading. The goal of stock market prediction is to forecast the future value of stocks as well as other financial instruments traded on stock exchanges. Most stock brokers base their stock predictions on technical, fundamental, or time series analysis. Python is the computer language used for machine learning techniques for stock market forecasts. In this research, we propose a Deep Learning approach that can be taught using publicly available stock data to learn from experience and then apply that information to make accurate predictions. A crucial area of research is stock price forecasting because of how profitable it can be for people, businesses, and governments. The method and novel applications are examined in this paper. This study examines novel applications and techniques for forecasting a particular corporation's consistent closing price. Customers can purchase or sell assets of the companies that are regularly scheduled there. A difficult task is predicting changes in stock market prices. In this study, stock price predictions are made utilizing deep learning models like the Long Short- Term Memory (LSTM), a development of the recurrent neural network. For the purpose of prediction, the two-year datasets from 2021/02/23 to 2022/07/20 are used. This study examines novel applications and techniques for forecasting a particular corporation's consistent closing price. Customers can purchase or sell assets of the companies that are regularly scheduled there. A difficult task is predicting changes in stock market prices. In this study, stock price predictions are made utilizing deep learning models like the Long Short- Term Memory (LSTM), a development of the recurrent neural network. For prediction, the two-year datasets from 2021/02/23 to 2022/07/20 are used.Item Quantum Walk-Enhanced Hybrid Routing: Integrating Local and Non-Local Best-Effort Strategies for Robust Quantum Networks(Comilla University, 1-Feb-2025) Dhruba, Anjan Das; Hasan , Md. Mehedinetwork nodes. Unlike classical networks, quantum networks require specialized routing strategie due to the no-cloning theorem, entanglement decay, and limited quantum memory. This paper investigates the performance of different distributed routing algorithms for entanglement distribution in a quantum internet. We analyze five routing algorithms, including modifiedItem Quantum Network Security :(Comilla University, 1-Feb-2025) Muhaimin, Md. Sayem; Mia, Md ShahinQuantum repeater networks are essential for long-distance quantum communication,