Akila de Silva

RipScout: Realtime ML-Assisted Rip Current Detection and Automated Data Collection using UAVs

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2025, Published Paper 

WIP: Citizen Science Tools with Machine Learning as a Pathway to Engage High School Students in Research

IEEE Frontiers in Education, 2024, Published Paper 

VortexViz: Finding Vortices by Learning from Particle Trajectories.

Under Review, 2024, ArXiv link

SmartCS: Enabling the creation of ML-Powered Computer Vision Mobile Apps for Citizen Science Applications without Coding.

Citizen Science : Theory and Practice, 2024, Published Paper 

Acceptance rate : 46%

Expanding partnerships, products and use-cases for an operational coastal web camera observation network

Ocean Sciences Meeting, 2024, link 

RipViz: Finding Rip Currents by Learning Pathline Behavior.

IEEE Transactions on Visualization and Computer Graphics (TVCG), 2023: Published Paper, Supp. Materials 

Acceptance rate : ~18-24%

How much does input data type impact final face model accuracy?

IEEE Computer Vision and Pattern Recognition (CVPR), 2022: Published Article  

Acceptance rate : 25.33%

Rip Current Detection from Region Based Convolutional Neural Networks.

Coastal Engineering, 2021: Video, Preprint, Published Journal Article, Data, BibTeX   

Acceptance rate : 27%

Authoring Platform for Mobile Citizen Science Apps with Client-side ML

CSCW Poster, 2021 : Poster, Published Paper 

Acceptance rate : ~ 60-70%

Face Models: How Good Does My Data Need To Be?.

IEEE International Conference on Image Processing (ICIP), 2021: Published Paper 

Acceptance rate : 46%

From Lifeguards to AI: Advancing Rip Current Forecasting Through New Detection Methods.

Ocean Sciences Meeting, 2020: Poster link, Poster

Sequential Image Synthesis for Human Activity Video Generation.

International Conference on Image Analysis and Recognition, 2019:  Published Paper

Acceptance rate : 59%