Presentations of virtual cultural heritage artifacts are often communicated via the medium of interactive digital storytelling. The synergy of a storied narrative embedded within a 3D virtual reconstruction context has high consumer appeal and edutainment value. We investigate if 360° videos presented through virtual reality further contribute to user immersion for the application of preserving intangible cultural heritage. A case study then analyzes whether conventional desktop media is significantly different from virtual reality as a medium for immersion in intangible heritage contexts. The case study describes bridge diving at Stari Most, the old bridge in Mostar Bosnia. This application aims to present and preserve the bridge diving tradition at this site. The project describes the site and history along with cultural connections, and a series of quiz questions are presented after viewing all of the materials. Successful completion of the quiz allows a user to participate in a virtual bridge dive. The subjective evaluation provided evidence to suggest that our method is successful in preserving intangible heritage and communicating ideas in key areas of concern for this heritage that can be used to develop a preservation framework in the future. It was also possible to conclude that experience within the virtual reality framework did not affect effort expectancy for the web application, but the same experience significantly influenced the performance expectancy construct.

The goal of psychometric scaling is the quantification of perceptual experiences, understanding the relationship between an external stimulus, the internal representation and the response. In this paper, we propose a probabilistic framework to fuse the outcome of different psychophysical experimental protocols, namely rating and pairwise comparisons experiments. Such a method can be used for merging existing datasets of subjective nature and for experiments in which both measurements are collected. We analyze and compare the outcomes of both types of experimental protocols in terms of time and accuracy in a set of simulations and experiments with benchmark and real-world image quality assessment datasets, showing the necessity of scaling and the advantages of each protocol and mixing. Although most of our examples focus on image quality assessment, our findings generalize to any other subjective quality-of-experience task.

Fast track article for IS&T International Symposium on Electronic Imaging 2018, Human Vision and Electronic Imaging 2018 Conference.

Teaching children with autism requires special set of tools and methods, due to decreased level of attention towards stimuli presented and lessened capability to learn in the ways typical children do, which is manifested within this population. It has been previously shown that computer-assisted intervention is not only an effective method for developing various skills, allowing both learning with teachers and practicing on their own time without the teacher's direct attention, but it nonetheless increases the motivation and results in faster acquisition of these skills.In this paper we present the first step in developing the LeFCA framework, that will be used for teaching children with autism basic skills and concepts. Within the pilot project, we produced four games for developing matching, pointing out (based on visual and auditory stimuli) and labeling skills, which are considered to be primary skills needed for learning. The results of our preliminary study showed that the created software in native language is completely clear and user friendly for kids with Autism and other special needs, and that is systematically and developmentally appropriately sequenced for learning. Additionally, we found that children were able to generalize learned skills, through a transfer to a new medium or environment without any needed training (i.e. computer). All four participants mastered all programs without any instructional tactic needed

Image compression standards rely on predictive coding, transform coding, quantization and entropy coding, in order to achieve high compression performance. Very recently, deep generative models have been used to optimize or replace some of these operations, with very promising results. However, so far no systematic and independent study of the coding performance of these algorithms has been carried out. In this paper, for the first time, we conduct a subjective evaluation of two recent deep-learning-based image compression algorithms, comparing them to JPEG 2000 and to the recent BPG image codec based on HEVC Intra. We found that compression approaches based on deep auto-encoders can achieve coding performance higher than JPEG 2000, and sometimes as good as BPG. We also show experimentally that the PSNR metric is to be avoided when evaluating the visual quality of deep-learning-based methods, as their artifacts have different characteristics from those of DCT or wavelet-based codecs. In particular, images compressed at low bitrate appear more natural than JPEG 2000 coded pictures, according to a no-reference naturalness measure. Our study indicates that deep generative models are likely to bring huge innovation into the video coding arena in the coming years.