Research Publications

Permanent URI for this communityhttps://rda.sliit.lk/handle/123456789/4194

This main community comprises five sub-communities, each representing the academic contribution made by SLIIT-affiliated personnel.

Browse

Filters

2019 - 201922020 - 20265

Advanced Search

Filter by

Settings

Subject: search.filters.subject.SVM

Search Results

Now showing 1 - 7 of 7
  • Thumbnail Image
    PublicationOpen Access
    Solar Hotspot Detection Using VHDL-Simulated Fixed-Point SVM: A Methodology Toward FPGA Realization
    (Faculty of Engineering, 2026-03) Fernando, N; Seneviratne, L; Weerasinghe, N; Rathnayake, N; Hoshino, Y
    Early detection of thermal hotspots in photovoltaic modules is critical to ensuring their efficiency, safety, and longevity. This study presents a complete end-to-end methodology for implementing a fixedpoint Medium Gaussian Support Vector Machine classifier using VHDL for a Field Programmable Logic Array. The approach begins with feature extraction from thermal images of healthy and defective solar panels, which focuses on MPEG-7 descriptors. The study shows that high impact for hotspot detection comes from blue chrominance contrast. A medium Gaussian SVM model is trained in MATLAB and converted to a fixed-point Q1.15 format for hardware compatibility. Key parameters, including support vectors, Lagrange multipliers, bias, and kernel scale, are extracted and verified in a custom Python environment to ensure numerical alignment with MATLAB results. The validated model is then implemented in synthesizable VHDL. It is verified using GHDL and the GNU Tool Kit waveform viewer, confirming bit-accurate hardware behaviour. Results show classification accuracy exceeding 99.3% with negligible performance loss due to quantization. The design achieves deterministic latency through an FSM-based structure and parallel feature processing for a 300-support vector and 222-feature system. This method enables low-power, real-time inference on a UAV-based edge platform, primarily focusing on drones.
  • Thumbnail Image
    PublicationEmbargo
    PRODEP: Smart Social Media Procrastination and Depression Tracker
    (Institute of Electrical and Electronics Engineers, 2022-11-04) Kulatilake, T.T; Liyanage, P.L.R.S.; Deemud, G.H.K.; De Silva, U.S.C; Sriyaratna, D; Kugathasan, A
    Procrastination refers to the voluntary delay of urgent tasks and can have several negative consequences such as stress, health issues and academic underachievement [47]. It is viewed within physiological research as a self-regulation failure [48]. Similar to procrastination, another severe problem which comes up within lots of people including students and teenagers is "Depression". Depression is a massively widespread problem among people around the world as well as in Sri Lanka [49]. As a result of procrastination and depression, students has to face academic underachievement. One of the main cause of these widespread problems are Social media over-usage [50]. Therefore this paper presents a new tracker which presented as a mobile application with four main components. This research study is about identifying and tracking users' facial emotions and eye-aspect ratio to analyze real emotions of the user via device inbuilt webcam to identify user fatigueness and procrastination. This study also analyzes user behavior in two selected social media platforms which are Facebook and Twitter and identifies the negativity and depressiveness of "Sinhala"content using Machine learning based Sentiment analysis approaches. Also as a companion, this paper introduces a chat-bot which communicates with the user in "Singlish"language. Our final products will be a complete mobile application which generates reports to the user based on the analysis done in the four components. As future work we will introduce AutoML approaches instead of traditional machine learning based approaches.
  • Thumbnail Image
    PublicationEmbargo
    Child Head Gesture Classification through Transformers
    (Institute of Electrical and Electronics Engineers Inc., 2022-11-04) Wedasingha, N; Samarasinghe, P; Singarathnam, D; Papandrea, M; Puiatti, A; Seneviratne, L
    This paper proposes a transformer network for head pose classification (HPC) which outperforms the existing SoA for HPC. This robust model is then extended to overcome the limited child data challenge by applying transfer learning resulting in an accuracy of 95.34% for child HPC in the wild.
  • Thumbnail Image
    PublicationEmbargo
    AuthDNA: An Adaptive Authentication Service for any Identity Server
    (2019 1st International Conference on Advancements in Computing (ICAC), SLIIT, 2019-12-05) De Silva, H.L.S.R.P.; Claude Wittebron, D.; Lahiru, A.M.R.; Madumadhavi, K.L.; Rupasinghe, L.; Abeywardena, K.Y.
    Adaptive authentication refers to the way that configures two factors or multi-factor authentication, based on the user’s risk profile. One of the most pressing concerns in modern days is the security of credentials. As a solution, developers have introduced the multifactor authentication. The multi-factor authentication has an adverse effect on user experience. This paper proposes a novel adaptive authentication mechanism which tries to eradicate the negative user experience of the traditional multi factor authentication systems. Adaptive authentication gathers information about each user and prevents fraudulent attempts by validating them against the created profiles. This approach will increase the usability, user-friendliness by introducing multi-factor authentication only when its necessary using a risk based adaptive approach. Furthermore, the solution ensures security by authenticating the legitimate user through collectively analyzing the properties, behavior, device and network related information. In the creation of the user profile, the adaptive authentication system will gather and analyze the user typing behaviors using a unique recurrent neural network algorithm named LSTMs with 95.55% accuracy and mouse behaviors using SVMs with 95.48% accuracy. In device-based authentication, a fingerprint is generated to the browser and to the mobile device which is utilized in the analysis of the accuracy rate of the authentication. Blacklisting and whitelisting of the networks and geo velocity of the authentication request are captured under the geolocation and network-based authentication. All the accuracy rates are fed to the risk-based authentication which helps the decision of re-authentication or in the grant of access to the system by analyzing the risk score generated for the authentication request.
  • Thumbnail Image
    PublicationEmbargo
    Melanoma Skin Cancer Detection Using Image Processing and Machine Learning Techniques
    (2020 2nd International Conference on Advancements in Computing (ICAC), SLIIT, 2020-12-10) Ahmed Thaajwer, M.A.; Ishanka, U.A.P.
    In humans, skin cancer is the most common and severe type of cancer. Melanoma is a deadly type of skin cancer. If it identifies early stages, it can be easily cured. The formal method for diagnosing melanoma detection is the biopsy method. This method can be a very painful one and a time-consuming process. This study gives a computer-aided detection system for the early identification of melanoma. In this study, image processing techniques and the Support vector machine (SVM) algorithms are used to introduce an efficient diagnosing system. The affected skin image is taken, and it sent under several pre-processing techniques for getting the enhanced image and smoothed image. Then the image is sent through the segmentation process using morphological and thresholding methods. Some essential texture, color and shape features of the skin images are extracted. Gray Level Co-occurrence Matrix (GLCM) methodology is used for extracting texture features. These extracted GLCM, color and shape features are given as input to the SVM classifier. It classifies the given image into malignant melanoma or benign melanoma. High accuracy of 83% is achieved when we combine and apply the shape, color and GLCM features to the classifier.
  • Thumbnail Image
    PublicationEmbargo
    DenGue CarB: Mosquito Identification and Classification using Machine Learning
    (2020 2nd International Conference on Advancements in Computing (ICAC), SLIIT, 2020-12-10) Mohommed, M.; Rajakaruna, P.; Kehelpannala, N.; Perera, A.; Abeysiri, L.
    This research paper discusses a web-based application that assists Public Health Officers in the dengue identification process. The mosquito classification is done using image processing and machine learning techniques. The training models are developed using Convolutional Neural Networks Algorithm, Support Vector Machine Algorithm, and K-Nearest Neighbors Algorithm to validate the results to determine the most accurate and suitable algorithm. this paper discusses the previous related research work on its significance and drawbacks while highlighting design, methods, and implementation in the solution. We conclude that the CNN algorithm provides the highest accuracy among the machine learning techniques used.
  • Thumbnail Image
    PublicationOpen Access
    Web Based Voice Controlled Advanced PACS to Diagnose Lungs Cancer and Related Anomalies
    (www.ijisrt.com, 2019-10) Ratnasingam, T; Sayanthan, A; Velummylum, E. S; Archchana, K; Pulasinghe, K
    - PACS (picture archiving and communication system) is a medicinal imaging technology that is utilized basically in medicinal services associations to safely store and dissect carefully transmit electronic pictures. PACS a needed asset in contemporary hospitals, has demonstrated its key position in the department of radiology for archiving and collecting medical images, followed by its inclusion with the department of radiology. In this paper we have included work 3D displaying of DICOM pictures, calculation of segmented cancer part with fine calculations, voice recognition for a program to get and translate correspondence or to comprehend and complete spoken directions, and forecast of malignancy utilizing examinations of cancer symptoms.