Dasith Edirisinghe



Machine Learning Engineer
Deep Learning Researcher

About

About Me



With a strong foundation in Computer Science, Mathematics, and Software Engineering, I am a Machine Learning Engineer and Researcher who recently graduated with first-class honors from the University of Moratuwa, Sri Lanka. My research interests center on developing advanced AI techniques that closely mimic human cognition, focusing on Computer Vision and Multimodal Representation Learning. Additionally, I am interested in optimizing AI models for efficient training and inference on specialized hardware.

News

News Highlights


  • Mar 2025 : 1 paper submitted to IEEE Access
  • Oct 2024 : 1 paper got accepted to appear at CLOUDCOM 2024
  • Dec 2023 : Graduated From University of Moratuwa
  • Sep 2022 : Successfully conclude GSOC 22' with Weaviate

Publications

Publications


  • Edirisinghe, D., Nimalsiri, W., Hennayake, M., Meedeniya, D., & Lim, G. Chest X-Ray Report Generation using Abnormality Guided Vision Language Model. Submitted to IEEE Access (Under Review)
  • Edirisinghe, D., Rajapakse, K., Abeysinghe, P., & Rathnayake, S. (2024). SpotKube: Cost-Optimal Microservices Deployment with Cluster Autoscaling and Spot Pricing. In Proceedings of the 2024 IEEE International Conference on Cloud Computing Technology and Science (CloudCom), pp. 87–94. IEEE. DOI: 10.1109/CloudCom62794.2024.00026

Education

Education


University of Moratuwa, Sri Lanka
Nov 2018 - July 2023

  • B.Sc. Engineering Honours - First Class
  • Specialization: Computer Science and Engineering

Experiences

Experiences

Working Experience

• Machine Learning Engineer July 2023 - Present
Rabot Inc, USA - Remote
Working as a Machine Learning Engineer and Researcher specializing in the Computer Vision domain

  • Optimized Vision-Language Models (VLMs): Fine-tuned VLMs for visual question answering and reasoning tasks, achieving a 30% accuracy improvement in a binary QA task through LoRA fine-tuning on A100 40 GB GPUs.
  • Explored Deep Contrastive Learning: Designed and implemented approaches for image similarity search using metric learning techniques (Siamese networks, proxy-anchor loss, and triplet loss) to improve accuracy and efficiency.
  • Hand Grasp Classification: Built a PyTorch and Mediapipe-based model to classify seven distinct hand grasps in packing stations with ~90% accuracy, leveraging advanced video analysis for operational efficiency.
  • Optimized YOLOv7 for Hailo Hardware: Improved YOLOv7 object detection inference pipeline through quantization, achieving real-time performance (30 FPS) while maintaining accuracy and confidence levels.
  • Developed Efficient ROS2 Inference Node: Created a ROS2 node using ONNX runtime, optimized for NVIDIA GPUs with CUDA and TensorRT, enabling efficient model inference in resource-constrained environments.

Technologies : PyTorch, VLMs(LLaVA, Llama), LLMs, Darknet, YOLO, ViT, CNN, ONNX Runtime, Hailo SW Suite, ROS2, GCP, Docker, Gitlab, Python,C++, Mender, CUDA, TensorRT


• Google Summer of Code Contributor May 2022 - Sep 2022
Weaviate, Netherlands - Remote
Final Report

  • During the summer of 2022, I worked under the mentorship of Weaviate engineers:
  • - Developed the text summarization module for Weaviate, which helps users to summarize search results.
    - Created the inferencing engine using FastAPI and the Hugging Face Transformers library.

Technologies : Weaviate, Python, Golang, Docker, FastAPI, HuggingFace


• Software Engineering Intern Dec 2021 - Sep 2022

  • I contributed to the Sysco Warehouse Management System (SWMS) under the SWMS NewUI team, focusing on modernizing the classic SWMS application:
  • - Fixed UI-related bugs using React JS.
    - Improved the service layer performance by re-implementing logic and optimizing queries using Java.

Technologies : ReactJs, Java, JavaScript, Git, PLSQL


• Software Engineering Intern April 2021 - Sep 2021

  • Developed a 3D image conversion system leveraging AWS lambda container image support.

Technologies : AWS lambda, AWS SQS, Docker, Python, JavaScript

Teaching Experience

• Teaching Assistant Aug 2022 - Dec 2022

CS3042: Database Systems, University of Moratuwa


• Teaching Assistant Aug 2022 - Dec 2022

CS3953: Technical Writing, University of Moratuwa


Tech Talks

• Deep Learning based Recommendation Systems February 2021

LiveRoom Tech Talks


• Software Engineering Best Practices September 2022

CSE, University of Moratuwa


• Google Summer of Code Awareness Session February 2023

CS & ES, University of Moratuwa

Projects

Projects



1. Representation Learning based Image Order Prediction University of Bern - Research Project

• This project involved training a sender-receiver system to encode and decode the order of image pairs using binary messages and learnable embeddings. A deep learning pipeline was formulated and implemented, integrating Sender and Receiver networks for effective image order encoding and decoding.
Learning Outcomes : Deep Learning Problem Formulation, Representation Learning, Loss function based optimization


2. Contrastive Learning External Project - Research Project

• This project involves researching and comparing various contrastive learning methods, such as SimSiam, BYOL, SWaV, and SimCLR. I implemented different contrastive model architectures for both training and inference, processing 1D ECG signals and enhancing them for downstream tasks like classification using PyTorch.
Learning Outcomes : Research and Development, Contrastive Learning, Self-supervised learning


3. AlphaGo University of Moratuwa - Semester Project

• This project is regarding the Classification of Anomalies in the Gastrointestinal Tract through Endoscopic Imagery. Deep Convolutional Neural Network with transfer learning was used. MobileNetV2 has used as the base model.
Learning Outcomes : CNN, Transfer Learning


4. Wildfire Predictor Hackathon - H2o.ai

• The goal of this project is to predict wildfires in Australia. I developed the predictive model using prophet
Learning Outcomes : Time Series Forecasting


4. iVoke University of Moratuwa - Semester Project

• iVoke focuses on detecting driver drowsiness by observing eye blink rate.
Learning Outcomes : Eye Blink Detection(OpenCV), Embedded Hardware(Rpi)

More projects on Github


Contact

Contact Me