HPC Projects

SL No	Name	College	Project Title	Project Description	Project Domain	Compilers Used
1	RINAS T NAZEER	Govt. College of Engineering, Kannur	A Comprehensive Study on Deep Learning-Based Approaches for Image Super-Resolution in Real-World Applications	The Proposed project aims to conduct a comprehensive analysis on state-of-the-art SR models using deep neural network architectures such as CNN, U-NET, and GAN. The project also aims to develop an advanced Super Resolution (SR) scheme using deep learning models that deliver improved efficiency and reduced computational complexity. This initiative addresses critical challenges in Super Resolution, such as blur and noise, while also enhancing spatial resolution.	Deep Learning	python, jupyter notebook
2	PARVATHY P CHANDRAN	Govt. College of Engineering, Kannur	Visual Forecasting	Visual time series forecasting is an innovative approach that transforms the traditional numerical time series forecasting problem into the computer vision domain. Here, time series data is represented as images, and a network is built to predict corresponding subsequent images. This method aims at explicitly forecasting time series data visually as plots or images, with deeplearning models, allowing for a more intuitive and insightful understanding of the future trends.	Deep Learning	Jupitor Notebook, python, Google Colab
3	JYOTHSNA S MOHAN	Govt. College of Engineering, Kannur	Malayalam text summarization	The proposed project aims to develop an advanced framework for abstractive summarization of Malayalam text using Large Language Models (LLMs). This endeavor holds significant potential to revolutionize the way complex Malayalam legal documents are interpreted and summarized, providing concise, coherent, and contextually accurate summaries. The project is expected to contribute to the field of natural language processing by advancing the capabilities of LLMs in handling low-resource and morpholog	Deep Learning	Anaconda
4	BABY SYLA L	College of Engineering Trivandrum	ENHANCING INDIAN SIGN LANGUAGE RECOGNITION USING DEEP LEARNING AND MUTLIMODAL DATA	This research aims to enhance Indian Sign Language (ISL) recognition and translation using advanced Deep learning techniques. By integrating multimodal data, addressing the linguistic structure of ISL, and achieving real-time processing, the proposed work seeks to overcome existing limitations.	Deep Learning	PYTHON
5	ANJU J S	College of Engineering Trivandrum	Emotion Recognition	Emotion recognition using deep learning involves using advanced artificial intelligence to understand human emotions from various cues like facial expressions, voice tones, written text, or even physiological signals. The goal is to create systems that can interpret how someone feels in a way that mimics human understanding.	Deep Learning	Python, Pytorch, Torch vision, Matlab
6	DEVIKA R G	College of Engineering Trivandrum	Glaucoma Detection	This project focuses on developing an automatic diagnostic system to aid clinicians in detecting glaucoma in retinal fundus images.	Deep Learning	Pytorch, TensorFlow, Keras, GPU
7	AKHILARAJ D	College of Engineering Trivandrum	Image Denoising	The Image Denoising project aims to enhance the clarity of digital images by removing noise while preserving important details. Using advanced techniques like wavelet transforms and deep learning, the research seeks to improve image quality for applications in fields such as medical imaging, security, and photography.	Deep Learning	Python, pytorch, torchvision, matlab
8	MILI MOHAN	College of Engineering Trivandrum	Stegomalware detection and analysis using Machine learning	The project works on the detection of steganographically obfuscated malware in images using machine learning techniques	Machine Learning	Tensorflow, pytorch, Jupyter Notebook, Keras,Cuckoo Sandbox, malware analysis tools, Steganographic tools
9	ANARGHA K	College of Engineering Trivandrum	Naturalistic Driving Action Recognition	Detection of different distracted human driving actions. And altering the driver for risky driver behaviour. Detecting and altering the driver distracted behaviour.	Machine Learning	Python
10	ARUNDHATHI V	College of Engineering Trivandrum	Anomaly detection in completed goods	Anomaly refers to deviation from normal conditions or a discrepancy that is not considered satisfactory. Anomalies can result in system downtime, decreased efficiency and higher maintenance expenses. Anomaly detection helps to identify abnormal or unexpected patterns in finished goods that vary from normal processes. This project aims to identify anomalies in completed goods to reduce the risk of malfunctioning.	Machine Learning	Python
11	SANDHYA L S	College of Engineering Trivandrum	Medical Image forgery detection	Aim of the project is to find a novel method for forgery detection in medical images.	Deep Learning	Tensorflow, pytorch, keras, jupyter notebook, computer vision tools
12	VIDYA P V	College of Engineering Trivandrum	Large Language Model for Malayalam	The aim of the project is to implement a Large Language Model for Malayalam language so that various NLP tasks can be performed on it.	Others	Keras, Tensorflow, PyTorch, TSNE
12	GEORGE THOMAS	College of Engineering Trivandrum	Computer Vision	sparse graph based clustering	Deep Learning	Scikit learn,Tensorflow,PyTorch ,PyTorch Geometric ,DGL Deep Graph Library,C++,Python,
13	VINITHA V	College of Engineering Trivandrum	IoT Security Using Machine Learning Algorithms	It includes detection and mitigation of IoT attacks and threats using machine learning algorithms	Machine Learning	pytorch, tensor flow federated, keras, flower framework for federated learning, FedML, tools for drawing, PySyft
14	RINI VIJAYAN	College of Engineering Trivandrum	Biometric Security using Machine Learning	To identify the real and fake samples used in biometric authentication based on machine learning algorithms.	Machine Learning	Tensorflow, Pytorch, Keras, Python
15	SREEDIVYA R S	College of Engineering Trivandrum	Emotion detection	Deep learning approach for emotion detection from facial images and wearable sensor outputs.	Deep Learning	Python, TensorFlow, PyTorch
16	VINYA VIJAYAN	College of Engineering Trivandrum	Machine Learning	Retinal Image Analysis using Machine Learning	Machine Learning	Pytorch
17	DIVYA PRASAD K H	Govt. Engineering College, Painavu	Computer Vision	Enhancement of images in adversarial conditions	Deep Learning	PyTorch, TensorFlow, OpenCV, Visual Studio Code, CUDA, python, NVIDIA GPUs + driver
18	SUDHEER TM	Government Engineering College Thrissur	Optimizing Privacy and accuracy in federated learning	Working with a large image dataset to work to identify the potential threats and privacy concerns in federated learning	Machine Learning	Tensorflow_federated, tensorflow_privacy, flower
19	SHIJIN KNOX G U	Govt. Engineering College, Sreekrishnapuram	Signal Processing	Alzheimer’s disease is a weakening neurodegenerative condition with profound cognitive implications, making early and accurate detection crucial for effective treatment. In recent years, machine learning, particularly deep learning, has shown significant promise in detecting mild cognitive impairment to Alzheimer’s disease conversion. This project synthesizes research on deep learning approaches for predicting conversion from mild cognitive impairment to Alzheimer’s disease.	Deep Learning	Python, Matlab, Pytorch, Torchvision
20	DEEPA S S	College of Engineering Trivandrum	Characterization and Localization of Cytogenetic and Molecular Aberrations for subtyping Acute Lymphoblastic Leukemia	Research Work	Machine Learning	Python Framework
21	SAKHI S ANAND	Govt. College of Engineering, Kannur	Authorship attribution of source code	The detection of AI-generated code involves working with large models, extensive code datasets, and complex evaluation pipelines, all of which demand significant compute. High Performance Computing enables efficient large-scale model inference, rapid data processing, and distributed hyperparameter tuning, which are critical for achieving high accuracy and robustness. HPC infrastructuresupports reproducibility and scalability, making it essential for building AI code detection systems.	Deep Learning	Pytorch with CUDA
22	DHANESH S P	Govt. Engineering College, Wayanad	Human Detection	Human Detection.	Machine Learning	To train a human detection model (e.g., YOLOv8 or custom Python 3.8, PyTorch 1.8, Ultralytics, NumPy, Pillow, PyYAML, OpenCV, CUDA 11.7 , cuDNN
23	ASWATHY M C	Rajiv Gandhi Institute of Technology, Kottayam	Spam Email detection Clustering and Classification, using different machine learning algorithms.	The goal is to classify SMS messages as spam or ham (not spam) using clustering methods — both on plaintext data and on encrypted data using homomorphic encryption (via TenSEAL).	Deep Learning	Python(using numpy, pandas, scikit-learn, tenSeal, Concrete-ml etc)
24	SABINA M A	Rajiv Gandhi Institute of Technology, Kottayam	federated learning	a decentralized machine learning in which a model is trained across edge devices (such as mobile phones) containing disparate datasets without sharing them	Deep Learning	fed ml, tensorflow federated
25	KAILAS NATH V D	College of Engineering Trivandrum	Dementia Detection from Multimodal Biomedical Signals	To develop a robust, accurate, and affordable machine learning and deep learning models for the early detection of Alzheimer’s disease (AD) by analyzing multimodal data, including imaging, clinical, and biomarker information.	Machine Learning	Python
26	PARVATHY A L	College of Engineering Trivandrum	WEATHER ADAPTIVE MODEL FOR AUTONOMOUS DRIVING USING METAHEURISTIC ALGORITHMS	The project aims at improving the detection performance of Yolo in adverse weather conditions by making use of metaheuristic optimization algorithms. This work demonstrates the significance of hyperparameter optimization in enhancing Yolo’s performance especially in imaging tasks. A comparison on detection performance of Yolov9 with and without different metaheuristic algorithms are evaluated here for adverse weather conditions.	Deep Learning	Python IDE, Computer vision, Deep learning frameworks
27	ROHITH R D	College of Engineering Trivandrum	Malayalam Chatbot for Government Service Query Reasolution	Malayalam Chatbot for Government Service Query Reasolution	Deep Learning	pytorch, tensorflow, CUDA
28	NOUMIDA A	College of Engineering Trivandrum	Analysis of Bird Vocalization using Deep Learning and Computational Modeling	The project includes identifying and separating bird species from field recordings using deep learning techniques.	Deep Learning	Python, Pytorch, Keras and deep learning tools
29	MALA J B	College of Engineering Trivandrum	Speaker Diarization and Content Analysis	Research aims to do speech processing and content analysis of the speech	Deep Learning	python and deep learning associated tools
30	JINUMOL K M	Govt. Engineering College, Kozhikode	A Unified Model for Non-Destructive Detection of Strawberry Quality Using HSI System and Transformers	This project aims to develop a non-destructive method for assessing strawberry quality using a combination of Hyperspectral Imaging (HSI) and Transformer-based deep learning models	Deep Learning	Python ,CUDA ,TensorFlow,Keras,Matlab
31	SHIBIL AHAMMED M	College of Engineering Trivandrum	Image Caption Generator	It is a deep learning model used to predict captions for an image. In this we use cnn for extracting features from input image and lstm for caption generation. It uses flicker8k dataset.	Deep Learning	Google Colab
32	SNEHA P M	College of Engineering Trivandrum	DEEP LEARNING COVID-19 FEATURES ON CHEST X-RAYS	To develop a deep learning-based system for automated classification of lung segmented images into COVID-19 cases or normal cases and to compare the performance of three different deep learning architectures (DenseNet121, InceptionV3, and VGG16) for this classification task.	Deep Learning	Jupyter Notebook/Anaconda
33	ARUNIMA TK	College of Engineering Trivandrum	Helmet detection and numberplate detection using YOLOv4 Algorithm	A realtime Helmet Detection and License Plate Detection system using YOLOv4 a state of the art object detection model and Googles Open Image Dataset v4. Darknet is used to training YOLO models. The project was completed in two parts in two datasets. Pretrained weights such as those obtained from darknet can be used to initialize the backbone network.	Deep Learning	Google Colab
34	SREEHARI V	College of Engineering Trivandrum	LLM based Question Answering System for Malayalam Medium Curriculum: A GraphRAG Approach	This project focuses on the development of a GraphRAG-based Large Language Model (LLM) Question Answering (QA) agent designed to cater to the Malayalam medium curriculum.	Deep Learning	TensorFlow, PyTorch, LangChain, Hugging Face, Neo4j,etc…
35	SHIBU KUMAR K B	Rajiv Gandhi Institute of Technology, Kottayam	ML	Face detection and lip reading	Machine Learning	Python with anaconda, Pytorch, tensorflow
36	FATHIMATHUL NEFALA	College of Engineering Trivandrum	A Multimodal Visual Question Answering System for Malayalam Language	Visual Question Answering is a technology that allows computers to answer questions about images. While many VQA systems exist, they mostly work in global languages like English, leaving speakers of regional languages such as Malayalam with limited access to this technology. This project focuses on creating a VQA system that can understand and respond to questions in Malayalam. A key part of this project is creating a dataset which includes both images and corresponding questions and answer	Deep Learning	python
37	ARUNIMA CV	Rajiv Gandhi Institute of Technology, Kottayam	Few shot learning	Few shot learning in Biomedical engineering field.	Deep Learning	Python
38	HIBA FATHIMA K P	Govt. Engineering College, Wayanad	CREDIT CARD FRAUD DETECTION BASED ON VARIATIONAL AUTOENCODER GENERATIVE ADVERSARIAL NETWORK	The Credit Card Fraud Detection System using VAE-GAN is a full-stack web application designed to detect fraudulent transactions in real-time. It integrates a React.js frontend for user interaction, a Flask backend for API handling, and a TensorFlow-based VAE-GAN model for anomaly detection. The system processes transaction data using a pretrained VAE-GAN model, which learns normal transaction patterns and identifies anomalies as potential fraud.	Machine Learning	Visual Studio Code
39	ANUGRAHA P P	Govt. Engineering College, Wayanad	Advancing E-Commerce Authenticity:Deep Learning and Aspect Features Based Framework for Detecting False Reviews	today’s digital world, people rely heavily on online reviews before making a purchase. However, many companies manipulate ratings by posting fake positive reviews to attract customers. This project aims to develop a deep learning-based framework that can detect fake reviews using techniques like BERT, CNN, and LSTM.This will help ensure that online feedback remains trustworthy.	Deep Learning	Visual Studio, Google Collab, Kaggle
40	NISHY RESHMI S	College of Engineering Trivandrum	Textual Entailment Classification	Given a sentence pair , the project aims to classify into three classes namely entailment , contradiction and neutral.	Deep Learning	python, pytorch, tensorflow
41	JOBY N J	Government Engineering College Thrissur	number plate recognition for police department	traffic surveillance	Machine Learning	python,nodejs,mongodb
42	ADEEB P H	College of Engineering Trivandrum	Botnet DGA Domain Name Classification using Transformer Network	Domain generation algorithm is a technic used by the botnet infected systems to connect with there central server. So identifying and classifying the Domain Names generated by these DGA algorithms are beneficial in the case of cyber security	Deep Learning	Python, pytorch
43	AISWARYA BP	College of Engineering Trivandrum	EYE GAZE BASED VISUAL COMMUNICATION AAC APPLICATION	Eye gaze based AAC board communication web application controlled according to eye gaze .	Machine Learning	Jupyter Notebook
44	FIZA SOORAJ KHAN	Government Engineering College Thrissur	classification and dectection on thermal images using various models	This project focuses on object detection and classification in thermal images using state-of-the-art deep learning models: YOLOv8, DETR (DEtection TRansformer), and EfficientDet. Thermal imaging is crucial for applications like night-time traffic surveillance and smart city monitoring, where traditional RGB-based models may fail due to poor visibility. A custom thermal dataset was prepared in COCO format, with annotated classes such as cars, bikes, and trucks,etc.. about 7 classes.	Deep Learning	Python 3.8+,Ultralytics YOLOv8 ,Transformers (Hugging Face),COCO API (pycocotools),Google Colab
45	RIJIN AMINA P	Government Engineering College Thrissur	Automatic number plate detection and character recognition	This project focuses on automatic vehicle number plate detection and recognition using deep learning. It combines YOLOv8 for detecting number plates in real-world images and fine-tuned PaddleOCR for accurately recognizing the characters on the plates. A custom dataset of over 1000 labeled images was used to train the detection model, achieving high precision and accuracy. Detected plates were cropped and labeled manually for OCR fine-tuning, with a custom character dictionary created to handle a	Deep Learning	Google colab
46	SARIKA K T	Government Engineering College Thrissur	Signal processing approaches for quality/intelligibility improvement of Malayalam Dysarthric Speech for human/machine recognition	Signal processing methods such as spectral enhancement, formant correction and prosody modification can improve the quality and intelligibility of Malayalam dysarthric speech for both listeners and recognition systems. When combined with machine learning and deep learning approaches, these techniques enable effective classification, severity assessment and adaptive enhancement to support human communication and automatic speech recognition.	Deep Learning	Jupyter Notebook and Google Colab
47	RASHMI VR	Government Engineering College Thrissur	Multimodal Signal Processing	Multimodal signal processing deals with the integration and analysis of data from multiple modalities (e.g., audio, video, text, physiological signals, or sensor data) to extract meaningful information. Unlike unimodal approaches that rely on a single type of data, multimodal systems aim to capture complementary features from different sources, improving accuracy, robustness, and interpretability. Different signals are captured, preprocessed, and fused to perform tasks.	Deep Learning	Jupiter Notebook & Google Colab
48	RAMEEZ MOHAMMED A	College of Engineering Trivandrum	Hate Speech Detection	Training LLM models to detect Hate and Offensive speech in social media posts	Deep Learning	Python, Tensorflow, PyTorch, Keras CUDA, cuDNN
49	SANA S NAVAS	College of Engineering Trivandrum	Greenify CET	Greenify CET is a smart waste monitoring system that uses CCTV footage to detect if students are disposing waste in correct color-coded bins. Students receive reward or penalty points based on their actions , promoting responsible waste disposal through automated, tech driven supervision.	Deep Learning	Docker,OpenCV,Python,PyTorch,CUDA
50	CHAITHANYA K S	College of Engineering Trivandrum	Greenify CET	Greenify CET is a smart waste monitoring system that uses CCTV footage to detect if students are disposing waste in the correct color-coded bins. If a violation is detected, facial recognition powered by machine learning or deep learning is used to identify the individual. Students receive reward or penalty points based on their actions, promoting responsible waste disposal through automated, tech-driven supervision.	Deep Learning	Docker, Python (v3.8 or above), TensorFlow and/or PyTorch (latest stable versions), OpenCV, CUDA and cuDNN (compatible versions for GPU acceleration)
51	ANANDU P N	College of Engineering Trivandrum	Greenify CET	Greenify CET is a smart waste monitoring system that uses CCTV footage to detect if students are disposing waste in the correct color-coded bins. If a violation is detected, facial recognition powered by machine learning or deep learning is used to identify the individual. Students receive reward or penalty points based on their actions, promoting responsible waste disposal through automated, tech-driven supervision.	Deep Learning	Python,PyTorch, OpenCV,Cuda,Docker
53	ASWAN RAMESH	Govt. College of Engineering, Kannur	automatic navigation for vehicle	This project focuses on developing a stereo vision-based system for depth estimation using a pair of stereo images.These disparity maps are then converted into depth maps using known camera parameters such as focal length and baseline distance. The system allows users to measure distances between selected points in the scene and further integrates a obstacle detection mechanism and path finding .	Others	matlab, python, jupitor notebook , google colab
54	MUHAMMED MUHSIN O	Govt. College of Engineering, Kannur	robust face identification against disguise	Face recognition with disguise identifies individuals despite masks, glasses, makeup, or wigs. Traditional systems fail as disguises alter visible features. Advanced methods use deep learning, thermal imaging, and 3D mapping to capture stable traits like bone structure and eyes. This ensures reliable identification in surveillance, security, and law enforcement applications.	Deep Learning	python,vscode
55	SREELAKSHMI K V	Govt. College of Engineering, Kannur	3D reconstruction from 2D images	3D reconstruction from 2D images is a computer vision technique that builds three-dimensional models of objects or scenes using multiple 2D images captured from different viewpoints. It involves feature detection, matching, depth estimation, and surface reconstruction.	Deep Learning	Jupyter notebook,python,colab,vscode
56	ANUSREE C	Govt. Engineering College, Kozhikode	Bruise detection and classification in hyperspectral images of strawberry using Deep Learning techniques	Bruise detection in strawberries is a critical task for ensuring fruit quality, reducing post-harvest losses, and maintaining consumer satisfaction. Traditional visual inspection methods often fail to identify subtle bruises that are not apparent to the naked eye, leading to compromised quality control. Hyperspectral imaging (HSI), which integrates spectral and spatial information across a wide range of wavelengths, offers a non-destructive and highly sensitive approach for detecting such hidden	Deep Learning	Matlab, Python
57	BRINTA BOSCO	College of Engineering Trivandrum	Bird Sound Classification	The objective is to develop a highly accurate and robust model that is effective for few-shot learning, performs reliably across diverse acoustic environments, and provides transparent, interpretable results for scientific and ecological applications.	Machine Learning	Python, CUDA
58	VISMAYA S	Govt. Engineering College, Painavu	Adaptive Two-Site Enterprise Networking Lab: Ludus-Based CCNA/CCNP Simulations with Security Awareness	This project focuses on creating a Ludus-based simulation lab representing two separate enterprise sites connected over a WAN. Each site is designed with its own LAN topology, including routers, switches, VLANs, and routing protocols, allowing learners to practice inter-site connectivity, network configuration, and performance optimization.	Others	LUDUS
59	ATHUL ANOOP	College of Engineering Trivandrum	Enterprise Resource Planner – RPL	The project is an enterprise resource planner for Rehabilitation Plantations Limited. The project is being carried out as part of a tender awarded to the college and being implemented by students of the Computer Science Department as an internship	Others	Git, Docker, NodeJS
60	JOAQUIM IGNATIOUS MONTEIRO	College of Engineering Trivandrum	Study of various machine learning algorithms for path planning and navigation of robots	Study of various machine learning algorithms for path planning and navigation of robots	Machine learning	Pytorch
61	ANJALI M	Govt. College of Engineering, Kannur	Modeling Higher-Order RNA–Disease Mechanisms using Hypergraph Graph Neural Networks	To develop a Hypergraph Graph Neural Network (HGNN) framework that integrates multi-source biological data (miRNA–lncRNA, miRNA–circRNA, and miRNA–disease associations) to predict lncRNA–circRNA–disease relationships by modeling higher-order miRNA-mediated interactions as hyperedges for discovering potential regulatory modules.	Deep Learning	Python
62	AKSHAYA P	College of Engineering Trivandrum	EpiBrain: An Interpretable Multi-Modal Deep Learning Model Integrating DNA Methylation and MRI for Early Alzheimer’s Prediction and Risk Stratification	Alzheimer’s disease (AD) is a progressive neurodegenerative disorder where early prediction is crucial for timely intervention. Existing AI models predominantly rely on single-modal data, such as MRI-derived features or molecular measures, and often function as “black boxes,” lacking interpretability into the drivers of disease progression. This project introduces a novel interpretable multi-modal deep learning model that integrates longitudinal peripheral blood DNA methylation and raw MRI image	Deep Learning	NVCC (CUDA compiler)
63	NEETHU G	College of Engineering Trivandrum	Brain stroke classification using Vision Transformer	This project classifies brain MRI images into Normal, Ischemic, and Hemorrhagic categories using the Vision Transformer (ViT) model. ViT divides each image into small patches and processes them as sequences using self-attention to learn important spatial relationships. The dataset contains labeled CT images, and data augmentation is applied to increase data diversity and improve model performance. The model is trained and validated to accurately distinguish between different stroke types. Unlik	Machine Learning	Google colab
64	SMRITI GOVIND	College of Engineering Trivandrum	SuperResolution	Enhancement of smartphone images using superresolution	Deep Learning	Python
65	VIDYA P V	College of Engineering Trivandrum	Large Language Model for Malayalam	The aim of the project is to implement a Large Language Model for Malayalam language so that various NLP tasks can be performed on it.	Others	Keras, Tensorflow, PyTorch, TSNE
66	ANOOPA S	College of Engineering Trivandrum	Deep learning enabled crowdsourcing of moving IoT devices	The research aims to determine the best possible IoT service providers for a moving IoT user to provide WiFi hotspot service from location A to an unknown location B, by considering Quality of Service parameters.	Machine Learning	Tensorflow, Pytorch
67	AJU K TOLY	College of Engineering Trivandrum	Bilingual sentence embedding BERT models for malayalam sentence embedding	The project involves preparation and cleaning of the required datasets for the pre training of a bert based language agnostic model and evaluating its performance on downstream tasks.	Deep Learning	Python, pytorch, cuda
68	RIMSHID ALI	College of Engineering Trivandrum	AUTOMATIC MODULATION CLASSIFICATION USING DEEP LEARNING	CLASSIFICATION OF MODULATION TYPES FROM THE RECIEVED SIGNALS IN A RECIEVER USED FOR RF INTELLIGENCE.	Deep Learning	PYTHON, PYTORCH,TENSORFLOW,KERAS,NUMPY,PANDAS,JUPYTER NOTEBOOK
69	MAHEEN K	College of Engineering Trivandrum	Multimodal Deepfake Detection	A multimodal deepfake detection project	Machine Learning	Python Pytorch
70	SREEHARI SURESH	College of Engineering Trivandrum	Multimodal Deepfake Detection	Deepfake detection using video audio textual features	Deep Learning	Tensorflow Python cuDNN CUDA
71	KEZIN B WILSON	College of Engineering Trivandrum	Robust Automatic Number Plate Recognition in Adverse Conditions: An Integrated Framework for Low-Light, Fog, Rain and Night time.	Robust Automatic Number Plate Recognition in Adverse Conditions: An Integrated Framework for Low-Light, Fog, Rain and Night time.	Machine Learning	colab,jupiter notebook
72	SANGEERTHA SREEJITH	Govt. College of Engineering, Kannur	Multimodal Explainable AI for Recurrent Fragility Fracture Prediction Using Small LLM and RAG	This project aims to develop a lightweight, multimodal, and explainable prediction system that integrates clinical records, X- ray imaging, and external medical knowledge using small language models (LLMs) enhanced with Retrieval-Augmented Generation (RAG). The system will provide accurate and interpretable predictions for recurrent fragility fractures while being resource-efficient and deployable in real-world clinical environments.	Deep Learning	Python
73	SREELAKSHMI K V	Govt. College of Engineering, Kannur	3D reconstruction from 2D images	3D reconstruction from 2D images is a computer vision technique that builds three-dimensional models of objects or scenes using multiple 2D images captured from different viewpoints. It involves feature detection, matching, depth estimation, and surface reconstruction.	Deep Learning	Jupyter notebook,python,colab,vscode
74	AUGUSTINE FELIX JOSHY	Govt. College of Engineering, Kannur	Hetrogenous Graph convolutional neural network for lncrna-drug association	The project develops a Graph Neural Network (GNN)-based framework to predict lncRNA–drug associations using a heterogeneous biological network integrating multi-source data (lncRNA–disease, drug–target, lncRNA–miRNA, disease–drug). A heterogeneous GNN model (HGT/HAN) learns node-type–specific embeddings for link prediction. Predicted associations are validated biologically and compared with existing machine learning baselines for performance evaluation.	Deep Learning	python
75	RAJASREE R	College of Engineering Trivandrum	Human Action Recognition	Classification and prediction of human actions from videos.	Machine Learning	Python
76	VISAKH R	Government Engineering College, Barton Hill	Problem Statement To develop computationally adaptive models to improve the resilience and robustness of machine learning classifiers against adversarial attacks by exploring novel defense strategies capable of detecting and mitigating adversarial inputs effectively.	Deep Learning	Google Colab, R Studio
77	BRINTA BOSCO	College of Engineering Trivandrum	Bird Sound Classification	The objective is to develop a highly accurate and robust model that is effective for few-shot learning, performs reliably across diverse acoustic environments, and provides transparent, interpretable results for scientific and ecological applications.	Machine Learning	Python, CUDA
78	HIBA SENBEKH C	Govt. Engineering College, Kozhikode	PROTEIN LIGAND BINDING AFFINITY PREDICTION USING DEEP LEARNING	This project focuses on predicting the binding affinity between proteins and ligands using advanced deep learning techniques. Protein–ligand interactions play a crucial role in drug discovery, as they determine how strongly a potential drug molecule binds to its target protein. Traditional computational methods like molecular docking or molecular dynamics simulations are often time-consuming and computationally expensive. In this project, we are using a deep learning–based model.	Deep Learning	python
79	AVANI PRAMEEL	College of Engineering Trivandrum	BUILDING A TRANSPARENT CREDIT CARD FRAUD DETECTION SYSTREM USING EXPLAINABLE AI AND ADVERSARIAL DEFENSE FOR FINANCIAL CYBERSECURITY	The system will help financial institutions detect fraud more accurately while providing clear explanations for its decisions. By integrating XAI, it allows users and regulators to see why a transaction is flagged as suspicious, enhancing transparency. At the same time, adversarial defense techniques will make the system more resilient to manipulation by fraudsters. Through testing on real-world financial data will also be carried out.	Deep Learning	Visual Studio Code (VS Code), Python 3.x, Tensorflow, Anaconda
80	TOM SEBASTIAN	Govt. College of Engineering, Kannur	MediVault: Blockchain-Based Secure Health Record System for Hospitals	MediVault is a system where each hospital keeps its own private blockchain to store health records. To ensure that these private chains remain trustworthy, MediVault uses a process called anchoring. In anchoring, every hospital regularly sends the hash of its latest private block to a public blockchain. This hash acts as a permanent checkpoint. If anyone tries to change data inside a hospital’s private chain, the anchored hash on the public chain will no longer match, making the change immediate	Others	Docker, Docker Compose, Java JDK 21, Maven, Foundry