Faculty of Computing

Permanent URI for this communityhttps://rda.sliit.lk/handle/123456789/4202

Browse

Filters

2021 - 20253

Advanced Search

Filter by

Settings

Subject: search.filters.subject.Transfer learning

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    PublicationEmbargo
    Spatio-temporal graph neural network based child action recognition using data-efficient methods: A systematic analysis
    (Elsevier Inc, 2025-06-03) Mohottala, S; Gawesha, A; Kasthurirathna, D; Samarasinghe, P; Abhayaratne, C
    This paper presents implementations on child activity recognition (CAR) using spatial–temporal graph neural network (ST-GNN)-based deep learning models with the skeleton modality. Prior implementations in this domain have predominantly utilized CNN, LSTM, and other methods, despite the superior performance potential of graph neural networks. To the best of our knowledge, this study is the first to use an ST-GNN model for child activity recognition employing both in-the-lab, in-the-wild, and in-the-deployment skeleton data. To overcome the challenges posed by small publicly available child action datasets, transfer learning methods such as feature extraction and fine-tuning were applied to enhance model performance. As a principal contribution, we developed an ST-GNN-based skeleton modality model that, despite using a relatively small child action dataset, achieved superior performance (94.81%) compared to implementations trained on a significantly larger (x10) adult action dataset (90.6%) for a similar subset of actions. With ST-GCN-based feature extraction and fine-tuning methods, accuracy improved by 10%–40% compared to vanilla implementations, achieving a maximum accuracy of 94.81%. Additionally, implementations with other ST-GNN models demonstrated further accuracy improvements of 15%–45% over the ST-GCN baseline. The results on activity datasets empirically demonstrate that class diversity, dataset size, and careful selection of pre-training datasets significantly enhance accuracy. In-the-wild and in-the-deployment implementations confirm the real-world applicability of above approaches, with the ST-GNN model achieving 11 FPS on streaming data. Finally, preliminary evidence on the impact of graph expressivity and graph rewiring on accuracy of small dataset-based models is provided, outlining potential directions for future research. The codes are available at https://github.com/sankamohotttala/ST_GNN_HAR_DEML.
  • Thumbnail Image
    PublicationEmbargo
    Advancing Canine Health and Care: A Multifaceted Approach using Machine Learning
    (IEEE, 2023-06-26) Wimukthi, Y; Kottegoda, H; Andaraweera, D; Palihena, P; Fernando, H; Kasthurirathnae, D
    This research paper proposes a comprehensive approach to enhance the well-being of dogs through a range of innovative technologies. Firstly, we develop an automated system for dog breed and age identification using a Convolutional Neural Network (CNN) and a transfer learning model. This system aims to provide an efficient and reliable solution for dog owners and new adopters who are interested in discovering more about their canine companions. Secondly, we propose the development of a system that uses Reinforcement Learning to generate personalized meal plans based on a variety of factors such as the dog's breed, age, weight, health status, and emotional state. The system aims to provide dog owners with a reliable and effective tool for generating personalized meal plans that will enhance their pets' overall health and well-being. Thirdly, we present a dog disease recognition application that utilizes an artificial neural network (ANN) for identifying dog diseases based on their symptoms. Lastly, we introduce a real-time remote dog monitoring system using loT devices with edge computing to detect aggressive and anxious sounds. Our system provides an accurate classification of dog sounds related to aggression and anxiety, which can help dog owners detect and respond to potential issues early on. This research aims to provide dog owners and veterinarians with a range of technologies that can help them better understand and care for their furry friends.
  • Thumbnail Image
    PublicationEmbargo
    Deep learning approach to classify Tiger beetles of Sri Lanka
    (Elsevier, 2021-05-01) Abeywardhana, D. L; Dangalle, C. D; Nugaliyadde, A; Mallawarachchi, Y
    Deep learning has shown to achieve dramatic results in image classification tasks. However, deep learning models require large amounts of data to train. Most of the real-world datasets, generally insect classification data does not have large number of training dataset. These images have a large amount of noise and various differences. The paper proposes a novel architectural model which removes the background noise and classify the Tiger beetles. Here object location is identified using contours by converting the original coloured image to white on black background. Then the remaining background is eliminated using grabcut algorithm. Later the extracted images are classified using a modified SqueezeNet transfer learning model to identify the tiger beetle class up to genus level. Transfer learning models with fewer trainable parameters performed well than the total number of parameters in the original model. When evaluating results it was identified that by freezing uppermost layers of SqueezeNet model better accuracy can be gained while freezing lowermost layers will reduce the validation accuracy. The proposed model achieved more than 90% for the test set in 40 epochs using 701,481 trainable parameters by freezing the top 19 layers of the original model. Improving the pre-processing to localize insect has improved the accuracy.