http://dlib.iit.ac.lk/xmlui/handle/123456789/2306 2026-04-06T22:11:08Z 2026-04-06T22:11:08Z Sumanathilaka, T.G.D.K. Weerasinghe, Ruvan Priyadarshana, Y.H.P.P. http://dlib.iit.ac.lk/xmlui/handle/123456789/2272 2025-05-02T09:56:40Z 2023-01-01T00:00:00Z

Swa-Bhasha: Romanized Sinhala to Sinhala Reverse Transliteration using a Hybrid Approach Sumanathilaka, T.G.D.K.; Weerasinghe, Ruvan; Priyadarshana, Y.H.P.P. With the social and technological revolution, the usage of social media platforms and instant message services strengthens native language compatibility in the digital arena. The Sinhala and the Romanized Sinhala became the prominent typing languages among the general Sri Lankan community. Informal short-hand-based typing and short net acronyms were used for easier Sinhala typing. But Typing Romanized Sinhala using ad-hoc transliterations and getting the expected output in native Sinhala is less accurate and time-consuming. Therefore, this study aims to introduce a novel reverse transliterator which can back transliterate and suggest Romanized Sinhala to Sinhala words. The Transliterator has been modelled using the Statistical approach with Trigram and Rule-based model for back transliteration purposes and Knowledge-based with a Trie data structure for suggesting purposes. The proposed solution is capable of transliterating both formal and informal shorthand Romanized Sinhala. This hybrid model used in the study is capable of efficient transliteration with the word level accuracy of 0.84. This proposed model can be used in digital platforms to enhance the usability of native Sinhala communication in a much more efficient way.

2023-01-01T00:00:00Z Samarasinghe, Buddhin Saroj Lakmali, K.B.N. http://dlib.iit.ac.lk/xmlui/handle/123456789/2271 2025-05-02T09:59:29Z 2023-01-01T00:00:00Z

A Deep Ensemble Learning Approach for Venomous & Non-Venomous Serpents and Insect Species Identification Samarasinghe, Buddhin Saroj; Lakmali, K.B.N. In recent years snake and insect attacks have become a huge problem worldwide. Most species have similar colors and shapes, which makes it hard to tell them apart using typical techniques. Similarly, identifying different species of bees and wasps can be challenging, especially for non-experts in the field. Therefore, developing a reliable and effective method for recognizing these animals is essential to reducing the issues caused by snake and insect attacks and supporting wildlife conservation efforts. The proposed method in this research utilizes deep ensemble learning and transfer learning techniques to identify snakes, wasps, and bees accurately. Deep ensemble learning involves combining several machine learning models to make better predictions, and transfer learning is a technique that allows pre-trained models to be re-used for new tasks, saving time and computational resources. The proposed method utilizes deep ensemble learning and uses three base transfer learning models, DenseNet201, MobileN etV2, and InceptionResN etV2, for classification. The study obtained a training accuracy of 93%. The research classified 11 types of snakes, four types of bees, and four types of wasps, including all 19 classes. Moreover, the author developed a mobile application for user interaction and utilized Flask API for the logic tier. The mobile application enables users to take pictures or upload images of a snake, wasp, or bee and identify the species accurately. This feature makes the proposed method more accessible to people who encounter these animals daily. Overall, this research represents a promising step towards developing an accurate and effective system for identifying snakes, wasps, and bees, with practical implications for human safety and biodiversity conservation.

2023-01-01T00:00:00Z Kasif, Gibran Thondilege, Ganesha http://dlib.iit.ac.lk/xmlui/handle/123456789/2270 2025-05-02T10:00:33Z 2023-01-01T00:00:00Z

Exploring Music Similarity through Siamese CNNs using Triplet Loss on Music Samples Kasif, Gibran; Thondilege, Ganesha In the rapidly evolving digital music landscape, identifying similarities between musical pieces is essential to help musicians avoid unintended copyright infringement and maintain the originality of their work. However, detecting such similarities remains a complex and computationally challenging problem. A novel approach to address this issue is a song similarity detection system that utilizes a Siamese Convolutional Neural Network (CNN) with Triplet Loss for effective audio input comparison. The model is trained on a custom dataset from WhoSampled, an extensive database of information on sampled music, cover songs, and remixes. The dataset comprises pairs of audio samples and interpolations, making it suitable for the Siamese CNN approach. Incorporating Triplet Loss enhances the model’s performance by learning discriminative features for improved comparison. The performance of this system is assessed using a confidence interval-based metric, achieving a 96.86% accuracy at a 99.7% confidence level in determining the similarity between music samples. The solution provides a helpful tool for musicians to actively compare their creations with existing songs, helping to reduce the likelihood of unintentional plagiarism and possible legal issues.

2023-01-01T00:00:00Z Kariyawasam, Nimendra Somasundaram, Sharmilan http://dlib.iit.ac.lk/xmlui/handle/123456789/2269 2025-05-02T10:01:15Z 2023-01-01T00:00:00Z

ANNOE: Adaptive Neural Networks Optimization Engine Kariyawasam, Nimendra; Somasundaram, Sharmilan As the demand for deploying machine learning models on high-end mobile devices and IoT devices increases, the need for efficient machine learning model optimization becomes critical to execute on-device AI tasks. One of the primary challenges in this context is the retraining process of these models on IoT and mobile devices with new data, which is constrained due to limited processing power. This limitation hinders the generation of accurate domain-specific inference results, as models may not be as well-trained with real-world data, potentially reducing their predictive accuracy and generalizability when applied to practical scenarios.

2023-01-01T00:00:00Z