Paper: Neural network-based detection of virtual environment anomalies

Just had a paper Neural network-based detection of virtual environment anomalies accepted to Neural Computing and Applications, with Alfredo Nantes and Frederic Maire.  The paper is here at QUT ePrints, and here at Springer.

Abstract The increasingly widespread use of large scale 3D virtual environments has translated into an increasing effort required from designers, developers and testers. While considerable research has been conducted in assisting the design of virtual world content and mechanics, to date, only limited contributions have been made regarding the automatic testing of the underpinning graphics software and hardware. In this work, two novel neural network-based approaches are presented to predict the correct visualization of 3D content. Multi-Layer Perceptrons and Self Organizing Maps are trained to learn the normal geometric and color appearance of the objects from validated frames, and used to detect novel or anomalous renderings in new images. Our approach is general, for the appearance of the object is learned rather than explicitly represented. Experiments were conducted on a game engine to determine the applicability and effectiveness of our algorithms. The results show that the neural network technology can be effectively used to address the problem of automatic and reliable visual testing of 3D virtual environments.

Well done Alfredo!

Ross

Report: ThinkBalm 3D Virtual World Vendor Choice

ThinkBalm is well known in the Virtual World business for providing excellent business reports on this growing field. They have an excellent report here, that formalises the process of choosing 3D Virtual World vendor products.

Well worth the read, even if only to obtain an overview of vendor product functionality, especially in the conference, training and collaboration application domains.

Ross

Paper: APCCM 2010 Modelling in 3D Virtual Worlds

I have uploaded to QUT eprints the paper I presented at APCCM 2010 this week - Conceptual Modelling in 3D Virtual Worlds for Process Communication.

In this paper, I go into a little more detail about the 3D BPMN Editor I am developing in Open Simulator. I also discuss how to annotate a Virtual World with a Business Process Model (BPMN) in order to provide a place for a business analyst and a domain expert to combine their information to develop the conceptual process model for the business.

One comment from the audience was quite profound. In my approach, the Business Analyst brings the conceptual process model, and the domain expert brings the model of reality - buildings, layout and other artifacts representing the business. Together, these can be used to validate the correctness of the presented process model. Indeed this is done every time a consultant talks to a domain specialist and builds a model. However, when using the virtual world as a modelling space, the domain specialist provides (virtually) tangible information about the Enterprise.

Ross

Think Balm Immersive Software Analysis

Interesting article at ThinkBalm on some predictions for Immersive Software in 2010.

It is a worthwhile read with some good insights. However, the article seems to focus in on the collaboration and meeting capabilities of such spaces. While this is a rational argument, due to the quick return on investment, and the ability to easily use the spaces for communication work, I can't help thinking that there needs to be more exploitation of the design and prototyping capabilities of such immersive environments.

The same modelling capabilities that give you rooms, avatars and other devices in world, can give you very powerful tools for actual prototyping of products, both at the artifact and service level. My research seeks to leverage such environments for process modelling, in order to aid the process of validating the "prototype processes" with clients, via as-is and to-be model visualisation. Again, the tools will have to approach high levels of maturity in order to be useful to analysts, and to give a good return on investment.

My prediction is that such collaborative environments will be more and more used for product development, by geographically distant teams. Maybe not this year, but I expect that the ability to remotely develop artifacts collaboratively, in addition to remote conferencing, will be very useful to companies wanting to cut down on travel expenses, and to effectively utilise distributed expertise in their businesses.

Ross

Virtual Worlds Shakeout

Forterra systems, a major player in the Virtual World space, is now having some troubles, layoffs included.

It has been commented by some that these companies burnt up loads of venture capital, and then proceeded to struggle, with their business models struggling in these times of economic difficulties.

For my two cents worth, this has been expected anyway. Second Life is still a major leader in this space (discounting the WoW and Eve online games and others), as it seems to have an economic critical mass to survive the effects of the GFC, for the moment.

So, what will be the business models of the future that will survive? Aside from pure business factors, my gut instinct tells me that maybe it is to do with the server client model and the clunky nature of some of the content creation tools. In my research, the first thing that tends to be shot back at me by skeptics (of my BPM VE ideas) is the level of work required to set these simulations up. Server systems are laggy, containing complex 3D interfaces that are clunky and hard to learn, and it all just seems too complex to engage with as a business person with little background in 3D environments or games.

My usual response back to these points are the following:

1. We need to create tools that work more like a word processor, and less like 3DS Max. People are used to word processing simple documents, not detailed work. Word processors years ago were beasts to use, drawn from document formatting systems like TeX. Now we routinely use word processors with much better interfaces, to easily create documents for our own business and personal needs.

2. We need to treat worlds as a document, not as a cloud service per se. Sure, run the world in a cloud when you need to, but allow the world to be down loadable for localised interaction and secure/lag free collaboration, just like I do with a word processed document when I work with my colleagues. This is that way we all tend to collaborate, at the moment.

3. We need to implement the levels of quality control that the games companies use. Remember, if the world stutters at all, then the immersive experience is broken, and will put people off using the environment. The iPhone is popular, in part, due to its smooth interface. People's aggression to software is roughly proportional to K x Lag in the user interface, where K is a very large integer.

Possibly, when the user experience of such tools is much better, we will see greater uptake of such systems, and a better business climate for them.

Ross

Business Virtual Worlds - ProtoSphere

Much has been made of the generalised virtual worlds developed so far, including Second Life, Open Simulator, et al. Interesting now that a number of more specialised worlds are being developed for business collaboration purposes. ProtoSphere is one that I have come across here.

It provides sharepoint document integration (nice), but is still a server solution, so you have to open up your firewall for it to run. It reminds me somewhat of Google Lively for business applications.

Ross

Virtual Thin Clients

From the Yesha blog, a great link to an article about the use of LLMedia to show the output from a desktop application on a prim in SL.

This introduces all sorts of possibilities in the area of visualising running process applications, as you can show the application running in real time on a "terminal" in the Virtual Environment.

Now, if only we could open up the event streams in an OS to the Virtual Environment message passing scripts, then we could have some very interesting thin client, virtual application servers.

Ross

CFP: Massively Multiuser Virtual Environments (MMVE 2009)

CALL FOR PAPERS

The 2nd International Workshop on Massively Multiuser Virtual Environments (MMVE 2009) at the IEEE Conference on Virtual Reality 2009 (VR09)

March 15th, 2009
Lafayette, Louisiana, USA
http://peers-at-play.org/MMVE09/
--------------------------------------------------------

With the increasing number of potential users of virtual and augmented reality systems, as indicated by the recent booms of massively multiuser online societies, the design of distributed, massively multiuser virtual environments

(MMVEs) becomes increasingly important, posing new requirements on both distribution platforms and virtual reality systems. Facing these challenges is a community-spanning effort that requires pooling the resources and experiences of both the virtual reality and the networking/ distributed computing communities.

This workshop hopes to provide a link between these communities to foster the development of highly distributed, flexible and robust virtual environments. We aim to gather both practitioners and researchers under one roof to discuss their findings, incite collaborations, and move the state of the art forward.

Ross

Exit Reality

Have just got around to looking at Exit Reality, due to the fact that lots of people have asked me about it. I've missed the boat a little due to teaching commitments this semester taking up so much of my blogging time, but since it relates to my area, I thought a short review was in order.

My assessment in short, is that Exit Reality requires a lot more work to make it usable and attractive as an environment for socialising. It is trying to attempt something very difficult, the integration of web multimedia information and 3D interfaces. They should be congratulated on the attempt, but it just doesn't work for me.

Here are some reasons:

• Exit Reality is slow and buggy. Games and virtual environments require smooth reliable interfaces to make the paradigm work properly, especially as such bugs disrupt the immersive experience. The games industry understands this, that is why they all have such large QA sections. Exit Reality needs to invest in some more bug testing.
• The way they handle general web pages is problematic. My University home page came up as a sea of small signs that conceal each other, making it more difficult to traverse the website, thus undermining any benefit that 3D VEs can bring to web pages. I would have thought that some nice cone trees or hyperbolic views may have been better. Not sure they did much of a usability analysis before developing this general web component.
• Apartment style metaphors for socialising, while engaging in their novelty, are not that usable. It takes a number of seconds to traverse the space in 3D, that would otherwise require a couple of mouse clicks in a 2D page.

While I applaud an Australian company for sticking its neck out on a difficult technology, I am unfortunately critical of its present ability to be more useful than the present 2D social networking forms. I think they have to go back to the drawing board to rethink the idea.

Ross

2008 Australasian Virtual Worlds Workshop

Call for Participation

The Australasian Virtual Worlds Workshop is an event for Australasian researchers, educators and business people involved in virtual worlds, to meet and discuss topics related to virtual worlds.The aim of this workshop is to build local capacity and virtual world expertise that connects with global expertise.

This workshop builds upon foundations established by the Second Life Discovery Day held in 2007 at Monash University, Australia.

Proceedings from that workshop are available at the website www.avww.org.

Innnov8 by IBM

In November 2007, IBM launched a serious game for people to learn business process modelling called Innov8. This is another example of the avalanche of serious game applications that have been developed, exploiting the educational properties of games in areas ranging from the military, to emergency workers, and now to business process modellers. I will be downloading it, as soon as IBM authorise my application. I will then report on its value as an educational tool.

But until then, it looks like a first strike on the area of using virtual environments for business process management. The question is, what sort of tool support is available for devleoping such representations, and how effective is it really going to be as a modelling and educational tool. One could say that maybe the use of games as teaching tools is turning into the proverbial "when using hammer, every problem is a nail" scenario.

Ross