Research Publications

Now showing 1 - 2 of 2

Open Access
EuqAud: Detecting Gender Bias in Audio Datasets Using Polynomial Regression-Based Metric
(Institute of Electrical and Electronics Engineers Inc., 2026) Jayawardena, S; Haddela, P.S; Shyamalee, T; Ekanayake, A; Mudalige, T; Dhanawardhana, I
With the growing adoption of audio based AI systems in high-stakes domains such as healthcare, law enforcement, and social media, ensuring fairness particularly regarding gender bias has become critically important. While prior work on fairness has predominantly focused on disparities in model performance, bias inherent in training datasets remains underexplored. To address this gap, we propose EuqAud, a novel, pre-trained and traceable fairness metric that quantifies gender bias in audio datasets using raw acoustic features such as pitch, energy, amplitude, and voice activity. Unlike methods dependent on demographic labels such as race, age or language, EuqAud is designed to be demographic and language agnostic, enhancing its applicability across diverse contexts. The score is computed using an equation derived from polynomial regression with L2 regularization (Ridge regression), yielding robust and generalizable outputs. It spans a range from −10 to 10, where 0 denotes neutral, positive scores indicate male dominant bias, and negative scores reflect female dominant bias. For clarity, bias severity is categorized into three tiers: Neutral (EuqAud < 2), Moderate Bias (2 ≤ EuqAud ≤ 6), and Strong Bias (EuqAud > 6). Evaluation across multiple datasets demonstrates high predictive performance, with R2 values between 0.95 and 0.99. By focusing on dataset level bias rather than model outcomes, EuqAud offers a scalable and rigorous solution for advancing fairness in audio-based AI systems.
Embargo
Foody-Smart restaurant management and ordering system
(IEEE, 2018-12-06) Liyanage, V; Ekanayake, A; Premasiri, H; Munasinghe, P; Thelijjagoda, S
Customers play a vital role in the contemporary food industry when determining the quality of the restaurant and its food. Restaurants give considerable attention to customers’ feedback about their service, since the reputation of the business depends on it. Key factors of evaluating customer satisfaction are, being able to deliver the services effectively to lessen the time of consumption, as well as maintaining a high quality of service. In most cases of selecting a prominent restaurant, customers focus on their choice of favorite food in addition to available seating and space options. Long waiting times and serving the wrong order is a common mistake that happens in every restaurant that eventually leads to customer dissatisfaction. Objectives of this online application “Foody” is to address these deficiencies and provide efficient and accurate services to the customer, by providing unique menus to each customer considering their taste. This concept is implemented as a mobile application using latest IT concepts such as Business Intelligence, Data Mining, Predictive Analysis and Artificial Intelligence. This includes graphics and 3D modeling that provide existent physical information related to food such as colors, sizes and further user can view the ingredients of the meal as well as the available tables. In addition, the app shows the real-time map to the restaurant. Current table reservation status is indicated by the color change of the table. Unique food recommendation and it’s order for each customer is generated by analyzing their social media information and the system notifies the customer the wait time by calculating it. Preparation of food and allocation is done subjectively. The expected outcome of the research is to develop a fully automated restaurant management system with the mentioned features as well as to avoid confusions between orders, provide better view of food and allow the customer to choose the menu according to their taste in a minimum time.

Research Publications

Browse

Filters

Advanced Search

Filter by

Settings

Sort By

Results per page

Search Results