SIGMA360

Spherical & Gestures Media Archive – 360° Quality Study.

Measuring what truly matters in immersive 360° video - quality, comprehension, and inclusion.

Login to access repository
Latest news

Recent project updates

View all news
13.06.2026.

SIGMA360 Web Repository Application Successfully Developed

The SIGMA360 project team has completed the development of the project’s web repository application, which will be used to host 360° video materials and provide public download links for test video content, subjective quality ratings, and related research data. The repository is currently prepared as a technical platform, while video materials will be added after the recording and processing activities begin.

Read more
12.06.2026.

SIGMA360 Team Proposes Curriculum for a New Graduate Course

As part of the SIGMA360 project activities, the project team has prepared a curriculum proposal for a new graduate-level elective course titled Quality of Service Analysis. The course is designed for students of the Information and Communication Traffic programme and focuses on the analysis of service quality, user experience, multimedia services, and modern communication networks.

Read more
About SIGMA360

Understanding quality in immersive 360° video

SIGMA360 explores how users perceive, assess, and interact with immersive 360° video, with a strong focus on quality, accessibility, and reproducible research.

The project combines subjective quality studies, sign-language comprehensibility assessment, and a structured media repository designed to support future research in immersive media.

Explore targets

360° video quality

Subjective assessment of immersive video quality and user experience.

Accessibility

Evaluation of sign-language comprehensibility in immersive media scenarios.

Open repository

Structured archive for video content, metadata, quality ratings, and research data.

Funding institution
European Union
Programme
NextGenerationEU
Funding type
Competitive EU project
Project duration
10/2025 – 09/2029
Project type
Scientific Research Project
Amount
111.575,27 EUR

Targets and Key Results

Measurable outcomes SIGMA360 will deliver through open science, inclusive evaluation, and reproducible engineering.

Scientific output and dissemination
Targets
Scientific output and dissemination
Open resources
Targets
Open resources
Data collection
Targets
Data collection
Model performance
Targets
Model performance
Capacity building and knowledge transfer
Targets
Capacity building and knowledge transfer

Select a target

Scientific output and dissemination

Peer-reviewed publications that validate methods, models, and insights from SIGMA360.

We will publish the project’s findings in high-quality open-access journals and leading international conferences, ensuring visibility, reproducibility, and meaningful comparison across the research community.

Targets
Q1/Q2 open-access papers: 3 Proceedings full papers: 8
Why it matters

Peer-reviewed dissemination makes the dataset, protocols, and models citable and comparable - accelerating progress in immersive VQA.

Open resources

Public deliverables aligned with open science and community engagement.

SIGMA360 will release the dataset and provide open web access to the repository, complemented by public VR demo events that showcase results and support adoption across academia, industry, and education.

Targets
Public dataset release: 1 Open web repository access: 1 Public VR demo days: 3
Why it matters

Open access enables reuse, benchmarking, and direct application in streaming, XR production, and assistive technology development.

Data collection

The core evidence base: clips, metadata, and subjective ratings at scale.

We will curate 360° clips with complete metadata and collect large-scale subjective ratings, including a dedicated sign-language comprehensibility track to support inclusive immersive services.

Targets
Clips with full metadata: 300 MOS ratings collected: 6000 CS ratings collected: 600
Why it matters

A content-rich dataset with MOS/CS labels allows robust training and testing of immersive VQA models under identical conditions.

Model performance

Quantified predictive accuracy on validation data.

SIGMA360 will deliver objective quality prediction models trained on the joint MOS/CS dataset and evaluated on unseen scenes, targeting reliable performance suitable for engineering workflows.

Targets
Validation target: RMSE ≤ 0.50
Why it matters

Accurate reference-free prediction supports adaptive streaming and network planning without costly new subjective tests.

Capacity building and knowledge transfer

Strengthening expertise, networks, and educational impact beyond the dataset.

The project invests in training and international exchange while translating results into practical guidance and teaching content, ensuring long-term value for researchers, practitioners, and students.

Targets
Accredited trainings completed: 2 Guest lectures hosted: 2 Guidelines for deaf users: 1 Graduate course proposal: 1
Why it matters

These outputs help turn research into practice - supporting inclusive XR services and modern teaching in multimedia quality engineering.

Workflow

From media preparation to validated research data

A reproducible, standards-aligned workflow that connects capture, controlled impairments, subjective evaluation, model validation, and open release in one consistent framework.

Designed for reproducibility, fair benchmarking, and accessibility-aware immersive media engineering.

Capture & metadata
Record diverse 360° scenes and compute SI/TI + descriptors.
Controlled impairments
Systematic sweeps of bitrate, fps, and resolution.
Subjective testing
ITU-T-style MOS plus sign-language Comprehensibility Score (CS).
Model training
Fuzzy / neuro-fuzzy prediction + validation on unseen scenes.
Open release
Web access, bulk download, DOIs, and permissive licensing.

People

The SIGMA360 team combines video engineering, human-factor testing, ML modeling, and sign-language expertise.

Assoc. Prof. Marko Matulin
Assoc. Prof. Marko Matulin

University of Zagreb

Faculty of Transport and Traffic Sciences

Leads the scientific and technical direction of SIGMA360, coordinating video creation, repository architecture, and objective quality modeling to ensure perceptual relevance and reproducibility.
Principal Investigator Repository & VQA models
Prof. Štefica Mrvelj
Prof. Štefica Mrvelj

University of Zagreb

Faculty of Transport and Traffic Sciences

Oversees ITU-T compliant subjective testing, dissemination activities, and financial management, ensuring methodological rigor and transparent project execution.
Subjective testing lead Dissemination & finance
Assoc. Prof. Ivan Grgurević
Assoc. Prof. Ivan Grgurević

University of Zagreb

Faculty of Transport and Traffic Sciences

Contributes expertise in risk, reliability, and preventive maintenance, integrating system robustness and operational realism into the project’s engineering workflows.
Risk & reliability Preventive maintenance
Assoc. Prof. Marko Periša
Assoc. Prof. Marko Periša

University of Zagreb

Faculty of Transport and Traffic Sciences

Leads accessibility evaluation and sign-language dataset development, connecting technical quality assessment with inclusive design requirements for deaf and hard-of-hearing users.
Accessibility lead Sign-language dataset
Petra Teskera
Petra Teskera, mag. ing. traff.

University of Zagreb

Faculty of Transport and Traffic Sciences

Provides data analytics support and bridges video-quality assessment with sign-language evaluation, aligning subjective data collection with analytical and modeling pipelines.
Data analytics support VQA ↔ HZJ liaison
Assoc. Prof. Marina Milković
Assoc. Prof. Marina Milković

University of Zagreb

Faculty of Education and Rehabilitation Sciences

Acts as an external expert for Croatian Sign Language, ensuring linguistic validity, ethical participant handling, and scientifically sound accessibility evaluation.
HZJ expert Recruitment & validation