Ranga Rodrigo
Ranga Rodrigo

Ranga Rodrigo, Ph.D.

Professor, Department of Electronic and Telecommunication Engineering, University of Moratuwa, Sri Lanka

I work in computer vision and deep learning. With my students and collaborators, I have worked on routing algorithms, improving convolutional layers, tracking, re-identification, few-shot learning, and contrastive learning, with applications in self-driving, robotics, and surveillance. We currently focus on predictive world models, vision-language navigation, multi-modal learning, and out-of-distribution generalization.

Research themes

All research
Figure for Vision-Language Models: Calibration and Adaptation

Vision-Language Models: Calibration and Adaptation

Making adapted vision-language models trustworthy: calibration through the angular geometry of text embeddings, and generalization to unseen domains.

  • M-PFT: Angular Margin for Calibrated Prompt-Tuning of Fine-Tuned Vision-Language Models WACV 2027 PDF Link
  • A-TPT: Angular Diversity Calibration Properties for Test-Time Prompt Tuning of Vision-Language Models ICLR 2026 PDF Code Link
  • SemAlign: Language Guided Semi-supervised Domain Generalization ICPR 2026 PDF
Figure for Computational Pathology

Computational Pathology

Explainable report generation from whole-slide images with knowledge graphs, fusion of histopathology with spatial transcriptomics, and label-efficient cancer subtyping.

  • PathKGen: Pathology Knowledge Assisted MLLM for Explainable Histopathology WACV 2027 PDF Code Link
  • SENCA-st: Integrating Spatial Transcriptomics and Histopathology with Cross Attention Shared Encoder for Region Identification in Cancer Pathology WACV 2026 PDF Link
  • Uncertainty Awareness Enables Efficient Labeling for Cancer Subtyping in Digital Pathology WACV 2025 PDF Link
Figure for 3D and 4D Scene Understanding

3D and 4D Scene Understanding

Representations for 3D and dynamic scenes: language-aligned 4D pre-training, generalized splatting, and learning on LiDAR point clouds.

  • CL4D: Contrastive Language-4D Pretraining for Vision-Language Reasoning in Dynamic Scenes ECCV 2026 PDF Code Link
  • DARB-Splatting: Generalizing Splatting with Decaying Anisotropic Radial Basis Functions WACV 2026 PDF Code Link
  • CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understanding CVPR 2022 PDF Code Link
Figure for Efficient Model Composition

Efficient Model Composition

Combining many LoRA adapters at once: a shared low-rank subspace computed once per pool gives composite quality at close to merge cost.

  • LoRASpace: A Pool-Wide Shared Substrate for Static and Dynamic Multi-LoRA Composition NeurIPS 2026 PDF Code Link
Figure for Vision for Self-Driving

Vision for Self-Driving

Perception for camera-based driving: the CeyMo road-marking benchmark and the lightweight SwiftLane lane detector.

  • SwiftLane: Towards Fast and Efficient Lane Detection ICMLA 2021 PDF Link
  • CeyMo: See More on Roads - A Novel Benchmark Dataset for Road Marking Detection WACV 2022 PDF Code Link

Recent publications

All publications
  1. Shihab Aaqil Ahamed, Mohamed Ansaff, Nang Khin Myat Lay Phyu, Udaya S.K.P. Miriya Thanthrige, Chamira Edussooriya, Ranga Rodrigo, and Muhammad Haris Khan, “M-PFT: Angular Margin for Calibrated Prompt-Tuning of Fine-Tuned Vision-Language Models,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2027. PDF Link
  2. Nadun Rajapaksha, Sandun Herath, Devnith Wijesinghe, Isiri Amani Withanawasam, Hasindri Watawana, and Ranga Rodrigo, “PathKGen: Pathology Knowledge Assisted MLLM for Explainable Histopathology,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2027. PDF Code Link
  3. Thamirawaran Sathiyalogeswaran, Navindu De Silva, Chamikara Siriwardane, W. K. Alwis, Jason Mars, Udaya S.K.P. Miriya Thanthrige, and Ranga Rodrigo, “LoRASpace: A Pool-Wide Shared Substrate for Static and Dynamic Multi-LoRA Composition,” in Advances in Neural Information Processing Systems, December 2026. PDF Code Link
  4. Kumal Hewagamage, Isuranga Senavirathne, Sasika Amarasinghe, Hasitha Gallella, Dulanga Weerakoon, Vigneshwaran Subbaraju, and Ranga Rodrigo, “CL4D: Contrastive Language-4D Pretraining for Vision-Language Reasoning in Dynamic Scenes,” in Proceedings of the European Conference on Computer Vision, 2026, pp. 594–612. PDF Code Link
  5. Muditha Fernando, Kajhanan Kailainathan, Krishnakanth Nagaratnam, Isuranga Senavirathne, and Ranga Rodrigo, “SemAlign: Language Guided Semi-supervised Domain Generalization,” in Proceedings of International Conference on Pattern Recognition, Leon, France, 2026. PDF
  6. Venuri Amarasinghe, Kalinga Bandara, Isun Randila, Asini Jayakody, Chamuditha Galappaththige, and Ranga Rodrigo, “Domain-Agnostic Feature Modulation for Semi-Supervised Domain Generalization,” in Proceedings of the IEEE/CVF Computer Vision and Pattern Recognition Workshop on Domain Generalization: Evolution, Breakthroughs, and Future Horizons, 2026, pp. 1–10. PDF

Meeting me

Either drop by or contact me over phone or email to make an appointment.

“Do not let your hearts be troubled.”

Resources

Links

Information for students: semester coordinators and others
Semester 1 (group coordinator):Dr. Rukshani Liyanaarachchi
Semester 2 (batch '20):Dr. Chamira Edussooriya
Semester 3 (batch '20):Prof. Tharaka Samarasinghe
Semester 4 (batch '19):Dr. Samiru Gayan
Semester 5 (batch '19):Dr. Upeka Premaratne
Semester 6 (batch '18):Dr. Subodha Charles
Semester 7 (batch '17):Dr. Peshala Jayasekara
Semester 8 (batch '17):Dr. Kasun Hemachandra
Mentoring:Dr. Samiru Gayan
Industrial training:Dr. Subodha Charles
Time table:Dr. Samiru Gayan
Biomedical program:Dr. Rukshani Liyanaarachchi
Research coordinator:Dr. Upeka Premaratne
FAC member:Dr. Chamira Edussooriya