Dr Tanmay Singha

PhD Computing, Curtin Uni (2023); MTech Info Tech, Uni Calcutta (2011); MCA, Uni Calcutta (2009)

Email: tanmay.singha@nd.edu.au

  • Biography

    Dr. Tanmay Singha completed his PhD in Computing at Curtin University (2020–2023), building on earlier postgraduate qualifications including an MCA (2006–2009) and an MTech in IT (2009–2011). With over 14 years of academic experience across India, Bhutan, and Australia, he has contributed significantly to curriculum development, student mentoring, and the delivery of high-quality computing education in diverse international contexts.

    His research expertise is centred on computer vision and machine learning, particularly object detection, semantic segmentation, and efficient AI models for resource-constrained devices. His research narrative is shaped by three pillars: (1) advancing visual recognition methods; (2) developing lightweight, deployable models for edge and embedded systems; and (3) applying AI solutions to support community, industry, and health-related needs.

    These research priorities align with UNDA’s themes of Health & Wellbeing—through AI-enabled support technologies—and Transformative Education & Innovation, reflecting his commitment to responsible, applied, and socially impactful computing research.

    In addition, Dr. Tanmay brings more than four years of industry experience as a software developer in Western Australia’s aged-care sector, enhancing his ability to integrate real-world practice with teaching and research.

  • Teaching areas

    Dr. Tanmay currently teaches core computer science programming and technical units, including Python, Java, Advanced Databases, and Computer Systems & Networks, with a strong emphasis on practical, hands-on learning. He has also taught a wide range of subjects such as Data Structures and Algorithms, Software Engineering, Requirements Engineering, and Web Application Development, allowing him to deliver both theoretical foundations and applied skills through project-based and interactive teaching approaches.

    With research expertise in artificial intelligence and computer vision, he is also interested in teaching advanced units in the future, including Artificial Intelligence, Machine Learning, Deep Learning, and Human–Computer Interaction.

  • Research expertise and supervision

    Dr Tanmay Singha's research expertise lies in artificial intelligence and computer vision, and he is available to co-supervise interdisciplinary Higher Degree Research projects that integrate AI-based techniques. Areas of interest include automated disease detection from images, image classification and segmentation, computer vision for healthcare, and applied machine learning solutions across different domains. To date he has co-supervised two Honours students on their final-year research projects, providing guidance on methodology, experimentation, and technical implementation.

    Prospective HDR students interested in these topics are welcome to contact Dr Signha for further discussion.

  • Journal articles and proceedings

    • Singha, T., Pham, D. S., & Krishna, A. (2023). A real-time semantic segmentation model using iteratively shared features in multiple sub-encoders. Pattern Recognition, 140, 109557. https://doi.org/10.1016/j.patcog.2023.109557
    • Singha, T., Pham, D. S., & Krishna, A. (2023). Multi-Encoder Context Aggregation Network for Structured and Unstructured Urban Street Scene Analysis. IEEE Access, 11, 66227-66244. doi: 10.1109/ACCESS.2023.3289968
    • Singha, T., Pham, D. S., & Krishna, A. (2023). Improved Short-term Dense Bottleneck network for efficient scene analysis. Computer Vision and Image Understanding, 235, 103795. https://doi.org/10.1016/j.cviu.2023.103795

  • Conference papers

    • Barker, M., Singha, T., Willans, M., Hackett, M., & Pham, D. S. (2025). A Domain-Adaptive Deep Learning Approach for Microplastic Classification. Microplastics, 4(4), 69.
    • Singha, T., Goswami, S., Pham, D. S., & Krishna, A. (2024, December). Lightweight Scene Parsing for Real-Time Structural Crack Detection. In International Conference on Recent Advances in Artificial Intelligence & Smart Applications (pp. 323-334). Singapore: Springer Nature Singapore.
    • Bevacqua, A., Singha, T., & Pham, D. S. (2024, November). Enhancing Semantic Segmentation with Synthetic Image Generation: A Novel Approach Using Stable Diffusion and ControlNet. In 2024 International Conference on Digital Image Computing: Techniques and Applications (DICTA) (pp. 685-692). IEEE.
    • Bergemann, M., Singha, T., Pham, D. S., & Krishna, A. (2024, November). Domain Adversarial SegFormer. In 2024 International Conference on Digital Image Computing: Techniques and Applications (DICTA) (pp. 677-684). IEEE.
    • Singha, T., Pham, D. S., & Krishna, A. (2023, November). Effi-Seg: Rethinking EfficientNet Architecture for Real-Time Semantic Segmentation. In International Conference on Neural Information Processing (pp. 55-68). Singapore: Springer Nature Singapore.
    • Singha, T., Pham, D. S., & Krishna, A. (2022, November). SDBNet: Lightweight real-time semantic segmentation using short-term dense bottleneck. In 2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA) (pp. 1-8). IEEE.
    • Singha, T., Bergemann, M., Pham, D. S., & Krishna, A. (2022, November). SC-CrackSeg: a real-time shared feature pyramid network for crack detection and segmentation. In 2022 international conference on digital image computing: techniques and applications (DICTA) (pp. 1-8). IEEE.
    • Singha, T., Pham, D. S., Krishna, A., & Gedeon, T. (2021, December). A lightweight multi-scale feature fusion network for real-time semantic segmentation. In International Conference on Neural Information Processing (pp. 193-205). Cham: Springer International Publishing.
    • Singha, T., Bergemann, M., Pham, D. S., & Krishna, A. (2021, November). SCMNet: Shared context mining network for real-time semantic segmentation. In 2021 Digital image computing: Techniques and Applications (DICTA) (pp. 1-8). IEEE.
    • Singha, T., Pham, D. S., Krishna, A., & Dunstan, J. (2020, November). Efficient segmentation pyramid network. In International Conference on Neural Information Processing (pp. 386-393). Cham: Springer International Publishing.
    • Singha, T., Pham, D. S., & Krishna, A. (2020, November). FANet: Feature aggregation network for semantic segmentation. In 2020 Digital Image Computing: Techniques and Applications (DICTA) (pp. 1-8). IEEE.
    • Singha, T., & Goswami, S. (2017, August). Classification in data mining using POEMs/GA algorithm. In 2017 International Conference on Energy, Communication, Data Analytics and Soft Computing (ICECDS) (pp. 3859-3864). IEEE.
    • Singha, T., Dutta, H. S., & De, M. (2012). Optimization of floor-planning using genetic algorithm. Procedia Technology, 4, 825-829.

  • Community engagement

    Dr. Tanmay has actively engaged with industry and the research community through a range of collaborative projects and professional contributions. His key projects include:


    1.Medical-Precision Facial Landmark Detection (Curtin School of Medicine)
    He completed a three-month internship with the Curtin School of Medicine, where he investigated deep neural models for medical-precision facial landmark detection. The project successfully replicated both 2D and 3D Face Alignment Network (FAN) models for jaw alignment and trained them on hundreds of thousands of images. The findings demonstrated strong model robustness to varying facial expressions and image resolutions, with the 3D FAN model achieving superior localization performance, particularly for non-frontal facial images.

    2.IRRIS Standalone Prototype V1.0 Analysis (Infinitive Group)
    In collaboration with the Infinitive Group, he evaluated the IRRIS Standalone Prototype V1.0, designed around a 60-GHz mmWave radar transceiver with three transmitters and four receivers. The investigation explored applications such as 3D people counting with fall detection, area scanning, and gesture recognition. The results highlighted the prototype’s potential for deployment in aged-care and hospital environments due to its capacity to detect falls and measure micro-movements related to breathing and heartbeat. However, the study also identified challenges, including signal-processing delays and noisy point clouds, recommending proper calibration before practical deployment.

    For both projects, a complete project report was formally prepared and submitted.

  • Affiliations and associations

    Dr. Tanmay has served as a reviewer for several journals and conferences, including Engineering Applications of Artificial Intelligence, the International Conference on Neural Information Processing, and the Australian Conference on Information Systems, contributing to research excellence and the dissemination of knowledge in computing and artificial intelligence.