Kanchana Ranasinghe

Avid problem solver, interested in the fields of Computer Vision and Machine Learning.

Complete resume for more details about me.

Education

PhD, Computer Science - CGPA: TBA
Stony Brook University, NY, USA

BSc (Hons), Electronics and Telecommunication Engineering - CGPA: 3.95 / 4.20
University of Moratuwa, Sri Lanka (Dean’s List: Semester 1, 2, 3, 4, 6, 7, 8)
Most Outstanding Graduand of the Year (valedictorian) - 2020

Work Experience

MBZUAI, UAE
Nov 2021 - Aug 2021: Research Assistant

VeracityAI, Colombo, Sri Lanka
Jan 2020 - Oct 2020: Machine Learning Engineer
Jan 2019 - Jan 2020: Associate Data Scientist

FiveAI, Cambridge, UK
June 2018 - Dec 2018: Research Intern

University of Moratuwa, Sri Lanka
July 2016 - Aug 2017: Undergraduate Researcher

Publications

Kanchana Ranasinghe, Muzammal Naseer, Munawar Hayat, Salman Khan, Fahad Shahbaz Khan, Orthogonal Projection Loss, ICCV 2021.
S. Ramasinghe, K Ranasinghe, Salman Khan, Nick Barnes, and Stephen Gould, Conditional Generative Modeling via Learning the Latent Space, ICLR 2021.
S. Jayasumana, K. Ranasinghe, M. Jayawardhana, S. Liyanaarachchi and H. Ranasinghe, Bipartite Conditional Random Fields for Panoptic Segmentation, BMVC 2020.
K. Ranasinghe, M. Jayawardhana, S. Liyanaarachchi and H. Ranasinghe, Extending Multi-Object Tracking systems to better exploit appearance and 3D information, 2019 arxiv preprint.
S. Ramasinghe, J. Rajasegaran, V. Jayasundara, K. Ranasinghe, R. Rodrigo and A. A. Pasqual, Combined Static and Motion Features for Deep-Networks Based Activity Recognition in Videos, in IEEE Transactions on Circuits and Systems for Video Technology, vol. 29, no. 9, pp. 2693-2707, Sept. 2019.
S. Ramasinghe, J. Rajasegaran, V. Jayasundara, K. Ranasinghe, R. Rodrigo and A. Pasqual, Micro Actions and Deep Static Features for Activity Recognition, 2017 International Conference on Digital Image Computing: Techniques and Applications (DICTA), Sydney, Australia, 2017, pp. 1-8.

Other Projects

Plant Disease Detection: Demo Video

Developed plant-leaf based disease detection system from multi-spectral image feeds (NIR/RGB spectra).
Deployment as android app and Raspi-based device

Realtime Multi-Object Tracking:

Combining Siamese Trackers and LSTM networks to simultaneously exploit appearance and spatial information
Evaluating effectiveness of BEV space projections for spatial tracking
Building system capable of operating in realtime

Programming Languages & Frameworks

Python: General algorithms and data structures, data pre-processing, visualizations, basic web development, and machine learning.

Python Frameworks: Tensorflow, PyTorch, Flask

Also possess basic knowledge in C++, C, Lua, and Verilog.