Digital avatars as humanized museum guides in the convergence of extended reality

Ana Martí Testón, Universitat Politècnica de Valencia, Spain, Adolfo Muñoz, Universitat Politècnica de València, Spain

Abstract

Keywords: Museum, Digital Avatar, Extended Reality, Museum Guide, Natural

Introduction

During the last ten years, we have experienced a revolution in the field of audio guides and information systems inside museums. These have gone from using museums’ analogue audio recordings to the deployment of users’ smartphones. The key to the development of these media guides is to provide relevant content to the visitor precisely when it is needed, mimicking the activity of ​​a guide who accompanies visitors during a tour and answers their questions (Winesmith, 2017). As a progression towards humanizing the experience of virtual guides, visual avatars are used as storytellers who facilitate content in a more affective and personal way, like real guides do in museums. Thanks to the latest technological developments, it is possible to create innovative storytelling centered on visitors’ experiences (Murawski, 2018) through interaction with avatars and digital content.

Figure 1: Virtual avatar seen from the point of view of a visitor wearing Hololens AR glasses.

A 3D virtual avatar is a computer representation of a human being, or an alter ego of a person, in the form of a three-dimensional model (Zhang & Yang, 2009). Avatars are normally used to humanize virtual experiences in video games and virtual tours. In these experiences, gamers assume an identity as an avatar and interact with other users (Suh, Kim, & Suh, 2011). Currently, many online communities are following the path that the virtual world “Second Life” initiated in 2003, where personal avatars play a central role in the way people represent themselves and communicate with other “players” (Kalning, 2011). Although “Second Life” is irrelevant nowadays, it paved the way to understanding the internet as a unique place to socialize with avatars, where the main idea of games is to communicate with other members and create stories.

Video games have dealt with avatars for more than 30 years (Mäyrä, 2008), and some important questions have arisen about their use in the museum context, such as the importance of user similarity to their avatars. Some publications demonstrate that the more closely an avatar resembles its user, the more positive the user’s attitude will be (e.g., affection, connection, and passion) toward the avatar, and the more prone the user will be to make a purchase (Suh, Kim, & Suh, 2011).

One of the main functions of museums is to disseminate their collections, attempting to enhance the relationship between objects and people (Witcomb, 2003, p. 127). This is where avatars can play an important role in creating a more empathic and moving experience. The latest technological developments of the 4.0 revolution are fueling museum practices with the use of Extended Reality (MX)– from Augmented Reality (AR) to Virtual Reality (VR) – Big Data (BD), and Artificial Intelligence (AI). The 4.0 revolution refers to the Fourth Industrial Revolution that uses artificial intelligence, big data, and algorithms with the massive interconnection of digital systems and devices. Extended Realty is a broad term that includes Augmented Reality (AR), as the incorporation of digital data and information in a real environment thanks to the recognition of patterns through software, and Virtual Reality (VR) as an interactive and immersive experience generated by a computer that replicates an environment, real or imagined, and simulates physical presence to allow interaction. Big Data is the set of data whose size, complexity, and speed of growth make it difficult to capture, manage, process, or analyze using conventional technologies and tools, and Artificial Intelligence (AI) is intelligence demonstrated by machines. In computing, an ideal “smart” machine would be a flexible rational agent that senses its environment and takes actions that maximize its chances of success in some goal or task.

As a consequence, new methodologies for the presentation of contents are emerging, creating original and sensitive narratives that engage in the traditional practice of storytelling (Gillam, 2017) and facilitate more immersive, intuitive, and smoother interaction between the virtual and the real worlds (Manovich, 2005, pp. 146-147) (Kipman, 2016) (Weiser, 1991).  All these principals can be applied to museology as a methodology to better face the new challenges of the 4.0 era, as was proposed by Dr. A. Martí in her thesis, directed by Dr. A. Muñoz. In that research, Museography 4.0 is defined as the set of techniques and practices related to museums which have evolved from analogical museography to the natural, immersive, and intuitive integration of digital data in the expositional context (Muñoz & Martí, 2020) (Martí, 2018) (Muñoz & Martí, 2018) while Museology 4.0 is the term that we propose to refer the philosophy, ideas, and concepts behind this new Museography. Thus, our main idea for Almoina AR was to design a virtual holographic guide as a storytelling prototype for an archeological museum. The Almoina Museum was the ideal place to present different virtual reconstructions of the Roman period of the city of Valencia as interactive layers with immersive information introduced by our digital guide, Cleia.

Human avatar representations can be divided in two groups: those that are very simple and cartoonish and those that try to copy the appearance and expressions of real people as closely as possible. The VR social application market is more based on personal avatars of the first type. An ecosystem of these applications has been developing since the successful appearance of the Oculus Quest 2 VR glasses, supported by Facebook. This standalone device has demonstrated the enormous interest the public has in avatars as a way of presenting themselves in online social events and experiences that range from watching a movie to visiting virtual worlds and sharing all types of digital assets.

Simultaneously, enterprises like Google and Nvidia have presented new hyperrealist 3D digital avatars prepared to speak different languages. These avatars will be integrated into a large variety of applications, from virtual desks in airports to virtual assistants in websites. The most impressive advance in the development of these prototypes is that some of them seemingly avoid the “uncanny valley” effect. This is the undesirable effect that 3D human representations produce when they are close to mimicking real humans, but do not completely succeed. Unfortunately, most of these technologies are still in the beta stage and cannot be tested in real scenarios.

The purpose of this study is to test a cost-effective way of producing AR avatars to create empathic experiences and engage museum audiences.

Our approach to the problem of human avatars specifically for museums is different from simplified versions and 3D models. It is based on video recordings of actors who play the roles of guides and historical characters at the same time. In this way, the “uncanny valley” effect is prevented. However, this format is less flexible to changes during production because all the necessary lines in the story must be recorded in advance, including any reaction or bifurcation in the plot that may be necessary as a response to specific interactions with visitors.

Related Research: The possibilities of Avatars in Museums

Museums have a wealth of information about the objects they keep, which must be organized and presented to appear attractive without overwhelming visitors. (Falk, 2009). Narratives associated with these objects must be adapted to this new media (Wong, 2015) to unite people and provoke dialogue that creates and transmits knowledge in an emotionally moving way (Parry, 2013). In fact, visitors’ experiences depend on a combination of factors (Hooper-Greenhill, 2016) (Bourdieu & Darbel, 2004). If museums want to be culturally relevant in today’s society, they need to take advantage of new technological resources (Witcom, 2016, pp. 580 -589) and play an active role in improving the museum experience (Falk, 2009). This can be done by developing strategies to rediscover heritage and reinvent its stories (Kenderdine, 2014). Thanks to digital media and the latest technological advances of the age of 4.0, museums have a new opportunity to disseminate collections which could be aided by virtual human avatars. In this context, digital museum guides and avatars are being used to mediate experiences with heritage both inside museums with Augmented Reality and on internet with Virtual Reality.

Multimedia guides have been demonstrated to contribute to a better understanding of museum contents (Eghbal-Azar, Merkt, Bahnmueller, & Schwan, 2015, pp. 133-142). They provide information but also engage visitors in an interactive exchange that can lead to deeper understanding and promote enthusiasm about content. Much of the technology available in the museum context has been used to create individual experiences (Sharples, FitzGerald, Mulholland, & Jones, 2013). However, new technological developments with Augmented Reality and Virtual Reality together with Artificial Intelligence and Big Data can facilitate social storytelling conducted by a virtual guide explaining content as a mediator, a friend, or an expert. Our current challenge is to design these experiences to be moving, social, and natural, mimicking human behavior during the museum visit and facilitating a content-driven experience (Kefi & Pallud, 2011).

The TOURBOT project (2001) was an early experience where a simplistic, humanized avatar robot was presented at the Hellenic Cosmos Cultural Center in Athens. Users tele-controlled the robot from home. They could scan the works and the avatar provided audio images and texts (Roussou, et al., 2001).

In 2009, it was difficult to interact with the avatars in multi-user virtual worlds. An early experiment created data-driven avatars that roamed a digital museum. The research was focused on three main technical issues: avatar animation, path planning, and collision detection among avatars (Wang, Meng, Yang, & Zhang, 2009). Other experiences proved that learning processes in museums could be more complete thanks to the use of avatars (Mu, Yang, & Zhang, 2009). In 2010, the Boston Museum of Science used two virtual avatars to prove that the presence of a virtual realistic actor can increase users’ engagement and attention in the museum context (Swartout, et al., 2010).

A recent article published in 2018 compares different storytelling approaches for virtual guides featuring virtual humans in a digital immersive museum. Some of the results demonstrate that, in terms of engagement and understanding of the content, users found it easier to visit the gallery through the avatar’s explanations, and this system helped to better focus their attention (Carrozzino, Colombo, TecchiaChiara, Evangelista, & Bergamasco, 2018).

The potential of virtual humans as storytellers can facilitate an affective narrative in virtual museums. A recent article analyzes three different avatars that personify a curator, a guard, and a visitor. Some of the findings of this research show that gender is an important factor when evaluating the effectiveness of the VR experience (Sylaiou, Kasapakis, Gavalas, & Dzardanova, 2020). The representations made with avatars bring up ethical issues that future museology must address. In our prototype, a white, brunette woman was selected to represent a credible Hispanic Roman character, but we are aware of the political and moral discussions that race and gender selection can have for audiences. Further investigation is necessary in this area.

Regarding the use of avatars inside galleries, the Museum of Celtic Heritage in Salzburg proposed an AR application for handheld devices, like tablets and smartphones, in 2016. A 3D virtual cartoon model of a Celt was used as a guide who tells stories about the museum. The app used Wikitude’s Augmented Reality SDK with Image Recognition and 2D Tracking technology (Breuss-Schneeweis, 2016).

AR applications can also use smart glasses to recreate Augmented Reality experiences inside museums. They are of special interest in historical and archaeological museums since they are capable of reconstructing contexts. However, they can also be used to present holographic avatars as mediators that explain museum content. Few articles about this topic have been published. One interesting example is the experience designed for the Egyptian Museum in Cairo. This research was focused on the idea of reducing the number of human tour guides in museums. It analyzed technical issues to permit the display of contents and interaction with digital data based on the immersion and presence theory, to be used with Head Mounted Displays like Hololens (Hammady, Ma, Strathern, & Mohamad, 2019). The results highlighted the importance of the mediating role guides play and the possibility of saving economic resources through the use of avatars (Hammady, Ma, AL‑Kalha, & Strathearn, 2021).

The creation of a holographic avatar for the Almoina Museum

In 2018, our research group developed a holographic video avatar prototype prepared to guide visitors through the Almoina archaeological museum in Valencia (Spain). The experience was designed to help visitors discover the meaning behind some cryptic ruins belonging to the Roman period, when the city of “Valentia” was created. The prototype was built for Hololens AR view-through glasses. The tour began with a mandatory introductory presentation lasting five minutes, where a holographic avatar named Cleia presented and explained the experience and how to interact with the holograms. Once the visitor learned how to operate the system, four interactive content presentations of between five and ten minutes were accessible, located in four important historical spots in the museum.

Figure 2: Holographic Production Schema.

The goals that directed this project were:

• To design and develop a functional prototype of an interactive human avatar as a virtual holographic presence in the museum.
• To decide what type of storytelling could meet the requirements of the Almoina archaeological museum. Since the Almoina museum holds different archaeological findings dating from the Roman period to the Middle Ages, it is difficult for a non-specialized visitor to discern between the dates of the different ruins in such a complicated layout.
• To apply the principles of Museology 4.0, which promote the use of technology to create more natural, interactive, immersive, intelligent, gamified, transmedia, and social experiences.

Methodology and production workflow

The design of the holographic experience had to be adjusted to meet the physical museography requirements. Museography 4.0 is a set of techniques related to museums adapted for our new digital era, using emerging, 21^st century technologies to achieve more natural, immersive, and intuitive integration of digital data in the exhibition context.

The following production scheme was applied when implementing the project:

Figure 3: Production Scheme for the Almoina Project.

The production of a multimedia, storytelling format using an avatar

The experience starts when visitors put on the glasses in the main area of the Roman ruins. They can hear Cleia calling them to come closer. The life-size video image of the avatar appears in a precise spot in the room. Once the visitor comes close enough, the system automatically triggers the start of the first presentation, where Cleia presents the experience and teaches the users how to select the interactive objects – red Roman Standards and green coins – that allow them to begin the tour. If a visitor does not activate these two items, the system will loop back to repeat each one until the visitor successfully completes the task. Once the introductory scene is finished, Cleia disappears from her initial spot and several red Roman Standards appear in different parts of the room, each one marking a new scene to discover with one main story again told by Cleia, and four coins floating in the space with extra information that can be activated in any order. Each one of these short stories is narrated by a male voice over a cinematic reconstruction of part of the ruins, lasting approximately 3-5 minutes. When the story ends, the Roman Standards and the coins that had momentarily been hidden, show up again, and visitors can replay any of the presentations if they wish.

The challenge of staging in the museum

The construction of Cleia as a flat avatar made from video feeds presented the challenge of creating a believable three-dimensional presence to visitors. The solution was planned from the very beginning of the project: the avatar would always stay at a minimum distance of two meters from the visitor and the plane of the video projection would be programmed to rotate in real time toward the person who was wearing the glasses. The hologram was also altered so that Cleia could make eye contact with museum visitors.

Thanks to the tracking features of the Hololens glasses, Cleia was programmed to appear in a spot of the Roman area beyond the railings, which was also beyond the reach of the audience and less than two meters away.

A young actress played the role of Cleia, the historic female Roman character acting as the visitors’ guide. It was necessary to shoot more than thirty videos per scene to collect all the reactions and sentences that the avatar must use to respond interactively to the visitors’ requests.

The video was shot on a chroma set in a small studio, while post-production was carried out with a compositing video editor to crop the video and create a video asset with a transparent background to be inserted into the application to become a hologram. The preservation of the transparency as an alpha channel was crucial to integrating 3D animated objects around Cleia in a realistic way.

To resolve the problem of the imprecise hand gesture input of Hololens version one, we decided to program input using only the position of the head to trigger the interactive red Roman Standard and green coins. The other common way of making selections is by voice commands, but this idea was rejected since museums are noisy places.

Figure 4: Videos of Cleia blended in the authoring tool with 3D objects in motion around her.

Figure 5: Animation of the interactive Roman Standards, activated by the visitor’s head.

Figure 6: Two moments of animations that rebuilt the Temple in the scene of the Sanctuary.

Programming was developed with the Unity video game engine. All the materials and assets, such as photos, videos, music, 3D objects, and animations were integrated to create the interactive scenes. After integrating all these assets and programming their “behavior”, the program was uploaded into the glasses to carry out the functional tests in real space.

The reconstructions of the museum space were intentionally programmed to appear and disappear gradually, so that visitors would have enough time to view how the virtual buildings fit into the layout of the real ruins. Special materials were programmed to make objects appear like progressive dissolutions of matter in space. This approach enhanced the magical aspect of the experience, giving it a feeling of traveling back in time.

A video recording of the experience watched from a visitor’s point of view can be seen at the following link:

https://figshare.com/s/3d95357310378a056608

Figure 7: Implementation of the Video Avatar inside the Almoina Musuem with Hololens glasses.

Integration and functionality tests and results

To check the feasibility of the application, different tests were carried out in February, 2018. Three formal sessions were conducted. The first one was the development of two Integration tests to check the proper functioning of the program, developed with ten people. Later, a usability test was carried out with 20 people to check the validation of the first alpha version with the introductory presentation and the one called “Santuario” (Sanctuary). In addition, an observational study was applied. Finally, a functional alpha version was generated integrating all the pieces: the video clips of Cleia.

The observational study showed that users stood an average of 37 minutes using the application, and 93% executed the training correctly. 90% decided to interact first with the red Roman Standards, and just 10% preferred to activate the green coins first. Most of them stood in front of the holograms and did not move around them. Most users initially reacted by trying to touch the holograms.

The users understood the tutorial that explained how to activate the contents very well, pointing their heads for two seconds toward the interactive signals. No one needed help, and they moved naturally around the room. Several of the users complained that there were big differences in volume between the guide’s voice (Cleia) and that of the other narrator. Possibly, some of the dialogues can be improved to make them shorter; more like open questions.

With the results of these usability tests, the team was satisfied. They had helped to verify that the proposed methodologies were functional, and that they met the proposed requirements. The methodology based on AR video avatars is viable and it has been demonstrated with these experiments.

Users highlighted the attractiveness of the application due to the novelty of the media. They also pointed out that it was very useful to understand the origins of the ruins and what constructions they belonged to. Moreover, they highlighted the ease of interacting with the virtual guide, the Roman Standards, and the coins.

Arguably, the main complaint made by the participants was that the field of view (FoV) of the device was too small (34 degrees diagonal). Hence, participants could see digital objects interacting with the real world while looking straight ahead, but if they turned their heads slightly, the digital objects disappeared or were cut off. It is important to point out that the users explained that once they were involved in the visit, they forgot about the field of view problem and became engaged in the storytelling.

Some visitors also reported discomfort caused by the weight of the glasses. Moreover, they pointed out the difficulty of wearing them with other glasses.

Conclusions and future applications

This application is an example of a new type of multimedia production able to create interactive human avatars without the disturbing 3D “realistic” model representations. The combination of an avatar presented as a guide and the creation of cinematic holographic movies that overlap real ruins with virtual content opens new opportunities for any kind of museum to create the kinds of experiences deemed “magical” by the testers.

For narrative, it represents a new way to discover hidden treasures from the past in places like archeological museums, where it is often difficult to truly value the remains exhibited.

Avatars will soon appear in smart phones and Extended Reality apps. In the near future, avatar-based solutions will become more cost-effective, combining smart phones with external AR glasses to help democratize the Museum 4.0 experience. However, some important questions remain for the future:

• How should a 21^st century avatar behave in a museum?
• Should we be able to choose the appearance of avatars to create more direct identification with personal cultural backgrounds?
• Should museums establish a closed criterion of appearance and behavior in accordance with their Museology plan?
• Should museums opt for hyperreal avatars to perfectly substitute a human guide or just play with simple cartoon representations?

This research explored a novel way of approaching stories related to heritage by testing a type of interactivity that facilitates the personalization of content according to visitors’ profiles and interests. This production system effectively applied some of the principles included in Museology 4.0. The use of an AR avatar with audio and animated 3D recreations has helped make the museum experience much more immersive, bringing the user closer to the feeling of taking a trip back in time.

Nevertheless, the introduction of Augmented Reality smart glasses in the museum context offers some challenges to be analyzed: In terms of production, the need for professionals with hybrid profiles as well as scientific 3D reconstructions will involve some additional media production costs. Concerning design, the dependence between the design of the interaction and the specific story that needs to be told should be mentioned. The high costs and maintenance of the devices must be considered. It can be concluded that specific methodologies will be feasible to use in museums in the near future, when AR glasses become more affordable for small institutions. In this regard, companies like the Korean Nreal are beginning to sell consumer-ready AR glasses which are much cheaper than the original Hololens, paving the way for a more competitive market in this area (Robertson, 2020).

In future research, the application will be reprogrammed to take advantage of the second version of Hololens that appeared in 2020, taking into account the wider field of view and the hand and eye tracking capabilities of the device.

References

Bourdieu, P., & Darbel, A. (1997). Love of Art: European Art Museums and their Public. Malden: Polity Press.

Breuss-Schneeweis, P. (2016). “The Speaking Celt – Augmented Reality Avatars Guide Through a Museum – Case Study. UbiComp ’16” in Pervasive and Ubiquitous Computing (eds.). Proceedings of the 2016 ACM International Joint Conference. Heidelberg, Germany: ACM. 1484-1491. doi: https://doi.org/10.1145/2968219.2974044

Carrozzino, M., Colombo, M., TecchiaChiara, F., Evangelista, & Bergamasco, M. (2018). “Comparing Different Storytelling Approaches for Virtual Guides in Digital Immersive Museums”. In C. Springer (Ed.), International Conference on Augmented Reality, Virtual Reality and Computer Graphics. 292-302. doi: https://doi.org/10.1007/978-3-319-95282-6_22

Eghbal-Azar, K., Merkt, M., Bahnmueller, J., & Schwan, S. (2015). “Use of digital guides in museum galleries: Determinants of information selection”. Computers in Human Behavior, 133-142.

Falk, J. H. (2009). Identity and the Museum Visitor Experience. Nueva York: Routledge.

Gillam, S. (2017). “Spotlight VR/AR: Innovation in transformative storytelling”. In MW17: Museums and the Web 2017. Cleveland Ohio. Retrieved February 11, 2018, from https://mw17.mwconf.org/paper/spotlight-vrar-innovation-in-transformative-storytelling/

Hammady, R., Ma, M., AL Kalha, Z., & Strathearn, C. (2021). “A framework for constructing and evaluating the role of MR as a holographic virtual guide in museums”. Virtual Reality. doi: https://doi.org/10.1007/s10055-020-00497-9

Hammady, R., Ma, M., Strathern, C., & Mohamad, M. (2019). “Design and development of a spatial mixed reality touring guide to the Egyptian museum”. Multimedia Tools and Applications, 79, 3465–3494. doi: https://doi.org/10.1007/s11042-019-08026-w

Hooper-Greenhill, E. (2016). “Changing Values in Art Museum. Rethinking Communication and Learning”. In B. M. Carbonell, Museum Studies. An athology of Contexts. Oxford: Blackwell Publishing Ltd. 517-532.

Kalning, K. (2011). If Second Life isn’t a game, what is it? Consulted January 15, 2021, from NBC News: Available https://www.nbcnews.com/id/wbna17538999#.U9uyeEi49yw

Kefi, H., & Pallud, J. (2011). “The role of technologies in cultural mediation in museums: anActor-Network Theory view applied in France”. Museum Management and Curatorship, 26(3), 273-289.

Kenderdine, S. (2014, abril 11). “How will museums of the future look?” In TEDxGateway 2013. Consulted January 10, 2016, from https://www.youtube.com/watch?v=VXhtwFCA_Kc

Kipman, A. (2016). “A futuristic vision of the age of holograms”. In TED talk. Consulted October 4, 2017, Consulted from: https://www.ted.com/talks/alex_kipman_the_dawn_of_the_age_of_holograms

Manovich, L. (2002). The language of new Media. Mit Press.

Martí Testón, A. (2018). “Hacia una museografía 4.0. Diseño de experiencias inmersivas con dispositivos de realidad aumentada”. Tesis Doctoral. Valencia: Universitat Politècnica de València. Available https://riunet.upv.es/handle/10251/107375

Mäyrä, F. (2008). An Introduction to Game Studies. London: SAGE Publications.

Mu, B., Yang, Y., & Zhang, J. (2009). “Implementation of the Interactive Gestures of Virtual Avatar Based on a Multi-user Virtual Learning Environment”. International Conference on Information Technology and Computer Science. IEEE Computer Society. 613- 617 doi: 10.1109/ITCS.2009.134

Muñoz, A., & Martí, A. (2018). “Holomuseum: A Hololens Application for Creating Extensible and Customizable Holographic Exhibitions”. EDULEARN18 Proceedings, 2303-2310.

Muñoz, A., & Martí, A. (2020). New Storytelling for Archaeological Museums Based on Augmented Reality Glasses. In S. Hageneuer, Communicating the Past in the Digital Age. London: Ubiquity Press. doi:https://doi.org/10.5334/bch.g

Murawski, M. (2018).”Towards a More Human-Centered Museum: Part 1, Rethinking Hierarchies”. Art Museum Teaching. Consulted January 31, 2018, https://artmuseumteaching.com/2018/01/22/rethinking-hierarchies/

Parry, R. (2013). “The Trusted Artifice, Reconnecting with the Museum’s Fictive Tradition Online”. In Drotner & Schrøder, Museum Communication and Social Media. Routledge Ltd.

Robertson, A. (2020). Nreal’s augmented reality glasses are shipping this month in Korea. Consulted January 20, 2021. https://www.theverge.com/2020/8/10/21362407/nreal-light-ar-glasses-lg-uplus-samsung-galaxy-note-retail-launch-availability-price

Roussou, M., Trahanias, P., Giannoulis, G., Kamarinos, G., Argyros, A., Tsakiris, D., Vassilis S. (2001). “Experiences from the Use of a Robotic Avatar in a Museum Setting”. Conference on Virtual Reality, Archeology, and Cultural Heritage. Glyfada, Greece. 153-160. Doi:  10.1145/584993.585017

Sharples, M., FitzGerald, E., Mulholland, P., & Jones, R. (2013). “Weaving Location and Narrative for Mobile Guides”. In K. C. Kirsten Drotner, Museum communication and social media the connected museum. New York. 177-196.

Suh, K., Kim, H., & Suh, E. K. (2011). “What If Your Avatar Looks Like You?” Dual-Congruity Perspectives for Avatar Use. MIS Quarterly, 35(3), 711-729. Doi/10.5555/2208923.2208935

Swartout, W., Traum, D., Artstein, R., Noren, D., Debevec, P., Bronnenkant, K., White, P. A.-Y. (2010). “Ada and Grace: Toward Realistic and Engaging Virtual Museum Guides”. International Conference on Intelligent Virtual Agents. Springer. 286-300.

Sylaiou, S., Kasapakis, V., Gavalas, D., & Dzardanova, E. (2020). “Avatars as storytellers: affective narratives in virtual museums”. Personal and Ubiquitous Computing, 24, 829–841. https://doi.org/10.1007/s00779-019-01358-2

Wang, X., Meng, X., Yang, C., & Zhang, J. (2009). “Data Driven Avatars Roaming in Digital”. The International Journal of Virtual Reality, 8(3), 13-18. Doi: https://doi.org/10.20870/IJVR.2009.8.3.2736

Weiser, M. (1991). “The computer for the 21st century”. Scientific American, 265(3), 94-104.

Winesmith, K. (2017). How SFMOMA made it’s audio-first, location aware app. MuseumNext. Melbourne. Consulted December 22, 2019. Available How SFMOMA Created it’s Location Aware Audio Tour – MuseumNext

Witcom, A. (2016). “Interactivity in museums. The Politics of Narrative Style”. In B. M. Carbonell, Museum Studies. An anthology of Contexts. Oxford: Blackwell Publishing Ltd, 580-589.

Witcomb, A. (2003). Re-Imagining the Museum. Beyond the Mausoleum. New York: Routledge.

Wong, A. (2015). “The whole story, and then some: ‘digital storytelling’ in evolving museum practice”. MW 2015 Museums and the Web 2015. Chicago: Museums and the Web. Consulted February 11, 2018. https://mw2015.museumsandtheweb.com/paper/the-whole-story-and-then-some-digital-storytelling-in-evolving-museum-practice/

Zhang, J., & Yang, Y. (2009). “Design and Implementation of Virtual Museum Based on Web3D”. In C. A. Pan Z., Transactions on Edutainment III. Berlin: Springer. Vol. 5940, 154-165 doi: https://doi.org/10.1007/978-3-642-11245-4_14