Lavanya Srinivasan

Dr Lavanya Srinivasan

Lecturer in Computer Science
School of Computing and Engineering
lavanya.srinivasan@uwl.ac.uk

Lavanya Srinivasan has been a lecturer in computer science at the School of Computing and Engineering since 2023. Lavanya teaches and leads various computer and information science modules, encompassing theoretical and practical dimensions across different academic levels. Her research focuses on image processing through computer vision techniques. Her goal is to develop innovative ideas that enable machines to learn from and comprehend the information conveyed in images. 

  • Qualifications

    • PhD, Image Processing
    • M.E, Systems Engineering
    • B.E, Electronics and Communication Engineering
    • Fellow of the Higher Education Academy 

Teaching

Currently teaching: Business Intelligence Technologies, Information Systems and Databases, and Introduction to Software Development. I am also keen on supervising postgraduate dissertation projects in image processing and computer vision, wishing to collaborate with motivated PG students to embark on their journeys into interesting topics of research.

Research

  • Research and publications

    • Lavanya Srinivasan, “Automated Recognition of Gait Emotions”, LNCS, Springer, 2024.
    • Lavanya Srinivasan, “Gender Identification Using 2D Ear Biometric”, Intelligent Computing, Springer, pp 439–449, 2024.
    • Lavanya Srinivasan. “Infallible Biometric Verification”, Computer Sciences, 2024.
    • Lavanya Srinivasan, “Automatic Sketch Generation for Person Recognition”, IEEE Access, pp.36-40, 2023. 
    • Lavanya Srinivasan, “Frequency Based Gait Gender Identification”, IEEE Access, pp. 13-17, 2023. 
    • Lavanya Srinivasan, “Automatic Gait Gender Classification Using Convolutional Neural Networks”, ACM International Conference on Image Processing and Machine Vision, pp. 34 – 40, 2023. 
    • Lavanya Srinivasan, “Analysis of Gait Biometric in the frequency domain”, IEEE Conference on Cloud Computing, Computer Vision and Image Processing, pp. 140-144, 2022.
    • Lavanya Srinivasan, “Identification and classification of COVID Images using Machine Learning”, International Conference on AI in Aging and Age-related Diseases, Elsevier, 2022.
    • Lavanya Srinivasan, “Gait Biometric for Person Re-Identification”, International Journal of Biomedical and Biological Engineering, Vol.15(12), pp.326 – 330, 2021.
    • Lavanya Srinivasan, “Inverse maximum likelihood-based edge detection for segmentation of breast lesion using active contour”, International Journal of Biomedical Engineering and Technology, Vol. 22 (3), pp.272-283, 2016.
    • Lavanya Srinivasan, “An approach to identify a lesion in infrared breast thermography images using segmentation and fractal analysis”, International Journal of Biomedical Engineering and Technology, Vol.19 (3), pp.220-229, 2015. 

  • Research degree supervision

    I am interested in supervising PhD students in the following areas: 

    Fall Detection and Prevention Using Machine Learning and Deep Learning

    An unexpected incident that forces people to take a seat on the lower level (floor or ground) is referred to as a fall. It consequently results in injuries that frequently have catastrophic consequences. Psychological complaints are also seen as a fall's aftermath. Anxiety, depression, activity restriction and a fear of falling are all possible mental health conditions. 

    The main physiological problem that older people face is a fear of falling, which limits their day-to-day activities. Because of their dread, older persons restrict their activities, which can impair their mobility and independence by weakening their muscles and causing inadequate gait balance. Understanding falls can be categorised into two categories for this purpose: fall detection and fall prevention. Fall detection is the process of identifying a fall by tracking a person's movements. Fall detection is the practice of spotting a fall by observing someone moving. Therefore, the research incorporates Machine Learning and Deep Learning for high classification accuracy for fall detection and prevention.

    Action recognition based on skeletons

    Identifying and characterising actions from a series of movements is the recognition task. Numerous disciplines, including computer science, psychology, and medicine, are connected to the study of human behaviour. Video surveillance, human-machine interaction, user interface design, gaming, entertainment, robotics, web-video explanation, medical diagnosis, sports analysis, and many more fields can benefit from the usage of action recognition.  

    There are numerous ways to perform an action, like jumping, based on the individual performing it, the situation, the surroundings, the movement style, and numerous other factors. Every activity can be performed in so many different ways that it is challenging to identify the characteristics that best describe it.

    Semantic segmentation

    An essential first step in understanding images is semantic segmentation. A dense classification problem is one in which each pixel in the image has a distinct label. Areas in pictures are frequently not grid-like and require information that is not local, which results in the failure of the classic CNN. It is natural to consider graph-structured data to handle it such as Graph-LSTM.

    Prospective PhD candidates may also suggest their own topics.

Courses taught on