Skip to content
@UM-SWIM-360

UM-SWIM-360

For Swimmers and Coaches: Swim-360 aims to help swimmers by combining video analysis and real-time biometric data for personalised insights.

UM-SWIM-360 · University of Malta

Deep Learning, Computer Vision & Explainable AI for Aquatic Sports Science

SWIM-360 Portal DIVE Portal University of Malta

Operating out of Room 0A19 at the University of Malta, UM-SWIM-360 is an elite research hub translating advanced AI architectures into track-side sports science realities. Directed by Prof. Vanessa Camilleri, our laboratory engineers high-frequency computer vision, multi-modal sensor fusion, and Explainable AI (XAI) systems capable of breaking down the complex physics of elite swimming.


🗂️ Explore the Research Architecture

Project Track Platform Access Core Foundations
SWIM-360 Ecosystem
Biometric & Video Fusion
🌐 swim-360.eu Explainable AI (XAI) Multi-Sensor Streams Data Dashboards
DIVE Ecosystem
Underwater Vision Pipelines
🌐 dive-project.eu Computer Vision Pose Estimation Kinematic Tracking

🛠️ Core Engineering Foundations

⚡ Multimodal Data Fusion

We ingest and temporal-sync disparate streams—matching high-speed underwater video matrix telemetry with live surface wearable sensors to chart metrics invisible to the naked human eye.

🔍 High-Fidelity Explainable AI (XAI)

No black boxes. We bake transparency constraints directly into our neural layers. When our models identify localized biomechanical inefficiencies or flag injury risks, they generate clear physical justifications for coaches and sports scientists.

🌊 Refraction-Resilient Pose Tracking

Sub-surface optics suffer from fluid turbulence, air bubbles, and extreme light refraction. Our models are actively fine-tuned to isolate and track anatomical joint chains through high-turbulency aquatic states.


🔬 Active Research Pipelines

🌀 SWIM-360: Holistic Swim Optimisation

  • Timeline: November 2024 — October 2026 Active
  • Objective: Synthesizing camera feeds with telemetry to optimize output mechanics while tracking real-time muscle-strain warning flags.
  • Funding: Jointly backed by Xjenza Malta and the Malta Digital Innovation Authority (MDIA) via the R&I Thematic Programmes (Digital Technologies).
  • Access Hub: swim-360.eu

📹 DIVE: Data-driven Intelligent Video Evaluation

  • Timeline: November 2024 — May 2026 Active
  • Objective: Developing standalone visual tracking intelligence to extract precise joint-angle kinematic graphs directly from unstructured video pools.
  • Funding: Financed by Xjenza Malta through the highly competitive Research Excellence Program.
  • Access Hub: dive-project.eu

👥 Laboratory Team

Academic Leadership & Core Investigators

Engineering & Research Staff

🧬 SWIM-360 Engineering Team 📹 DIVE Vision Team

⚠️ Data & Source Policy

The modules, datasets, and framework dependencies under this organization represent ongoing institutional research tracks. Public code access, pre-trained coordinate weight drops, and testing assets are published incrementally following strict anonymity validation to safeguard sensitive athlete biometric records and respect regional data-sharing guidelines.


University of Malta · Faculty of ICT · Room 0A19 · SWIM-360 & DIVE Research Labs

Popular repositories Loading

  1. .github .github Public

Repositories

Showing 1 of 1 repositories

Top languages

Loading…

Most used topics

Loading…