Day 1 - Monday 2nd of October
Lectures to be held in MAR 0.002 at TU Berlin
Visualization
08:50 – 09:00 Opening words, overview of the school training program, instructions about lab works, overview of social program, etc.
09:00 – 09:45 Lecture 1A (Robert Bregovic– Tampere University): Visual Perception: Vision and light, physiology of the human visual system, spatial vision, color vision, depth perception
09:45 – 09:50 Break
09:50 – 10:35 Lecture 1B (Robert Bregovic – Tampere University): Visual Perception – continued
10:35 – 11:00 Coffee Break
11:00 – 11:45 Lecture 2A (Sebastian Knorr – EAH Jena): Fundamentals in stereoscopic 3D: Stereopsis, depth budget, frame violation, native pixel parallax, 3D comfort zone, perceived depth
11:45 – 11:50 Break
11:50 – 12:35 Lecture 2B (Sebastian Knorr – EAH Jena): Stereoscopic 3D display technology: active and passive techniques, auto-stereoscopic and multiview displays, binocular rivalry issues
12:35 – 14:00 Lunch Break
14:00 – 14:45 Lecture 3A (Erdem Sahin– Tampere University): 3D and Light Field display technology: Vergence-accommodation-conflict, spatial multiplexing, time multiplexing, tensor display, holographic display of LFs, accommodation-invariant displays, perceptual modelling and design of LF displays
14:45 – 14:50 Break
14:50 – 15:35 Lecture 3B (Erdem Sahin– Tampere University): 3D and Light Field display technology – continued
15:35 – 16:00 Coffee Break
16:00 – 16:45 Lecture 4 (Tomas Sluka – CREAL): Advances in Sequential Near-Eye Light Field Displays: Sequential light-field display by CREAL creates high-fidelity 3D imagery with correct monocular depth cues. It eliminates focal rivalry and vergence-accommodation conflict, while it is highly light-efficient and combines with classical prescription lenses. Contrary to widespread notion, light-field can also be equally computationally and data-efficient as conventional stereo displays. This talk will introduce why and how.
Social programme: 18:30 City Boat Tour with Welcome Reception Dinner
Day 2 - Tuesday 3rd of October
Lectures to be held in MAR 0.002 at TU Berlin
Visualization & Quality Assessment
09:00 – 09:45 Lecture 1A (Elena Stoykova– Bulgarian Academy of Sciences): Holography: Historic overview of holography, wave optics, holographic displays (challenges and development), 3D capture and 3D content creation for holographic displays, computer generated holograms, self-interference digital holography, holographic printing
09:45 – 09:50 Break
09:50 – 10:35 Lecture 1B (Elena Stoykova– Bulgarian Academy of Sciences): Holography – continued
10:35 – 11:00 Coffee Break
11:00 – 11:45 Lecture 2A (Peter Lambert, Ghent University): Real-time 6DoF rendering for kernel-based light field representations: A high-level overview of the Graphics Rendering Pipeline; short introduction to kernel-based representations for Light Fields, in particular SMoE; ‘normal’/mathematical way of rendering images using kernel-based representation; how can the kernel-based representation be mapped to concepts of the Graphics Rendering Pipeline, in particular how real-time rendering can be achieved.
11:45 – 11:50 Break
11:50 – 12:35 Lecture 2B (Peter Lambert): Real-time 6DoF rendering for kernel-based light field representations – continued
12:35 – 14:00 Lunch Break
14:00 – 14:45 Lecture 3 (Kjell Brunnström, Emin Zerman – Mid Sweden University): Overview & Data Collection for Quantitative Evaluation: What is quality, QoE, QoE vs. UX; What are user studies? How to do user studies? Ethics, GDPR, etc. Informed consent; Scales (ACR, Likert, 1-100, pairwise comparisons, etc.), rating, ranking, quantifying quality; Experiment setup (participant selection, biases, training, selection of methodologies, pilot test); Screening of test persons (visual acuity, color blindness, etc.); Lab tests vs. crowdsourcing; Things to consider for immersive technologies (Motion sickness susceptibility, Simulator sickness).
14:45 – 14:50 Break
14:50 – 15:35 Lecture 4 (Kjell Brunnström, Emin Zerman – Mid Sweden University): Data Analysis for Quantitative Evaluation: Things to do after data collection — Outlier rejection, MLE soft rejection; Descriptive statistics: Mean opinion score, SOS (std of opinion scores); Regression – linear regression, logistic regression; Finding statistical significance.
15:35 – 16:00 Coffee Break
16:00 – 16:45 Lecture 5 (Kjell Brunnström, Emin Zerman – Mid Sweden University): Qualitative Evaluation Methods: Data collection (interviews, focus groups, observations, questionnaires etc.); How to select the right method?; Data analysis; Ethical considerations; Reporting findings.
Social programme: Attendance of the German Unity Day celebration
Day 3 - Wednesday 4th of October
Lectures to be held in MAR 0.002 at TU Berlin
Quality Assessment
09:00 – 09:45 Lecture 1 (Kjell Brunnström, Emin Zerman – Mid Sweden University): Combination of Qualitative and Quantitative Methods: Why do we need to combine qualitative and quantitative methods? Specific designs: convergent/parallel designs, explanatory sequential, exploratory sequential, transformative; When should mixed methods be used? Triangulation (using multiple data sources); Combining data; Analysis techniques; Reporting mixed findings.
09:45 – 09:50 Break
09:50 – 10:35 Lecture 2 (Kjell Brunnström, Emin Zerman – Mid Sweden University): Image/Video Quality Metrics: Automatic quality estimation introduction – Full reference, RR, NR; Image quality vs. video quality; Analysing automatic quality estimation results
10:35 – 11:00 Coffee Break
11:00 – 11:45 Exercises (Kjell Brunnström, Emin Zerman – Mid Sweden University): Exercises for automatic visual quality estimation and for data analysis with python/Matlab
11:45 – 11:50 Break
11:50 – 12:30 Wrap up, Discussion on TS outcomes
12:30 – 14:00 Lunch Break
14:00 – 15:30 Workshop opening words, Overview of planned work (15 minutes per group)
15:30 – 16:00 (Coffee Break)
15:30– 17:00 Setup and preparation of subjective tests
Social programme: 19:00 Paint Ball and Dinner
We meet Ernst-Reuter-Platz at 17:30 (precise location: Google maps)
Day 4 - Thursday 5th of October
The workshop will be held at the Communications Systems Department at TU Berlin (Building EN. 3rd floor)
09:00 – 10:30 Setup and preparation of subjective tests
10:30 – 11:00 (Coffee Break)
10:30 – 12:30 Preparation and execution of subjective tests
12:30 – 13:30 Lunch Break
13:30 – 15:00 Preparation and execution of subjective tests
15:00 – 15:30 (Coffee Break)
15:00 – 17:00 Evaluation of subjective tests
Social programme: 19:00 Dinner @ Solar Restaurant & Bar
Day 5 - Friday 6th of October
Lectures to be held in MAR 0.002 at TU Berlin. The workshop will be held at the Communications Systems Department at TU Berlin (Building EN. 3rd floor)
09:00 – 09:45 Lecture 1A (Peter Eisert – Fraunhofer HHI/ Humboldt University Berlin): Rendering and Image Synthesis: From Classical Approaches to Neural Methods: This short course addresses different aspects of rendering and image synthesis. We will start with classical render approaches with rasterization methods as well as methods for support of global illumination aspects (e.g. raytracing or radiosity). In recent years, the classical approaches have been extended by deep learning based techniques that support neural rendering of scenes with very high quality. The second part of the course will therefore dive in into novel neural rendering aspects and describe methods using learned texture descriptions or neural radience fields.
09:45 – 09:50 Break
09:50 – 10:35 Lecture 1B (Peter Eisert – Fraunhofer HHI/ Humboldt University Berlin): Rendering and Image Synthesis: From Classical Approaches to Neural Methods – continued
10:35 – 11:00 Coffee Break
11:00 – 13:00 Visit of the Innovation Center for Immersive Imaging Technologies & TiME Lab @ Fraunhofer HHI
13:00 – 14:30 Lunch Break
14:30 – 16:30 Presentation of results (10 minutes per group) and wrap up of workshop
Social programme: 19:00 Festival of Lights
