Creating the visual aspects of the visages that define us within virtual spaces, is no simple task. There are immense roadblocks, both computational and perceptual, in the way. It is not just a matter of creating a believable body, although that in itself is a considerable challenge. There is the additional challenge of animation and flow in realistic ways as well.
Sections
Hair is one of the central aspects of a person’s visual identity. It is sculpted and fashioned ten thousand different ways in self-expression, physically. Virtually, it is often a nightmare to reproduce. Still, if we are to have realistic, believable, enjoyable avatars, realistic hair is a must.
Creating skin that looks realistic is a major challenge. If it is smooth, flawless, it looks like a mannequin or a cartoon. Real skin is imperfect, pockmarked, discoloured. Many fine tweaks need to be made, to create skin that is believable and comfortable for the avatar’s user.
Processing an Avatar's Facial Colour
 Researchers at Toyohashi University of Technology have discovered that the human bain processes the colour of a face separately to the features of that face. This is an interesting development, especially when placed in the context of crafting personalised avatar forms for AI sales agents and other interactive AI in virtual space.
Back To Top
The face is the single most important aspect of self-presentation, non-verbal communication and self expression. This is where almost all the sensory organs are gathered physically, and where the most attention is paid by others. To create a believable face is a must for avatar communication of any real value.
ACQUINE and Virtual Identity
In a handful of the more pioneering virtual environments, a system called FaceGen, along with other, similar systems, allows a user to photograph their face from front and side, and use that to put together a 3D model of their physical head, if they so desire, to use for the basis of their avatar presence online. Other technologies are just coming into use, that allow adjustments, based on attractiveness, of that face.
Androgynous Avatars Inspire Less Trust
 Two researchers set out to see if mixed gender avatars would be less well received than single gender. Their findings? A person of androgynous appearance is felt to be less trustworthy than one clearly defined as belonging to one gender or another, when all other factors are eliminated.
A counterpart to the other facial expressions book by the same author, Babies to teens starts at birth and carries forwards to age 19, whereas the other book deals with ages 20 to 90.
Facial expressions, is exactly what you would expect. It is predominately a huge collection of physical human faces, ranged 20-90, photographed in a huge variety of expressions and emotional states.
Instantaneous Visual Virtual Face
 One of the major issues with any virtual form, is facial expression. Traditionally, getting a virtual face to match your physical intent for expression in real-time, was a lost cause. Even for big budget film making, CG overlays had to be constructed frame by frame by hand. Enormously time consuming, ludicrously expensive and completely useless for real-time usage.
An expressive face is a work of art. Constantly moving and changing. Lips, brows, frown lines, each is in constant motion. Stop Staring analyses facial structures and movements, then shows animators how to bring life to the faces of their characters.
The Uncanny Valley: Effect of Realism on the Impression of Artificial Human Faces
A MIT Presence magazine free feature. Roboticists believe that people will have an unpleasant impression of a humanoid robot that has an almost, but not perfectly, realistic human appearance. This is called the uncanny valley, and is not limited to robots, but is also applicable to any type of human-like object, such as dolls, masks, facial caricatures, avatars in virtual reality, and characters in computer graphics movies.
Video games need 'realism boost'
BBC article about how to add authenticity to VR, goes beyond graphics, also encompassing extensive use of motion capture to catalogue how stance, gait and the tiny movements of facial muscles combine when people display different emotions.
Back To Top
The eyes, windows onto the soul. Dead eyes aare an instant turn off, as are eyes that don’t move right, look off, or don’t reflect the environment. They are a hideously complex challenge, but an eye that looks wrong, makes the whole face look wrong.
Back To TopSkeletons and Boning (17)
Without a skeleton, it is very difficult for an avatar to move. Boning points are not actual bones, but hinge points between stiff sections. Without careful design and consideration, the joints will not move realistically. Without bones, they can still move, but in a way which never escapes the uncanny valley.
Eadweard Muybridge's pioneering work on the motion of all manner of quadrupeds. Horses, goats, cats, gnus, eagles, gazelles, sloths, camels, many others shown walking, running, flying, leaping, almost a complete range of natural motions duplicated for the benefit of animators and skeletal model making.
Bypassing the Uncanny Valley with Hands
 In August 2008, the first hand and forearm pairing was achieved which realistically bypasses the uncanny valley for the lower arm, and allows completely realistic movement of the wrist and hand.
Horses and other Animals in Motion is a collection of, as the title says, 45 sets of photographs of horses hauling, walking, trotting, etc., plus sequences of donkeys, an ox, pig, dog, cat, deer and other animals capture details of anatomy and movement. These images, were taken by the definitive expert in the field, Eadweard Muybridge.
The parts specific to clay working are irrelevant to the virtual developer, but the general techniques and practices are invaluable if you choose to take this route, and create anatomically believable avatars. Ideal for working with weighted bone animation systems.
This book and CD is essentially the electronic format version of Eadweard Muybridge?s ?Animals in Motion?. It contains electronic format versions of 167 black-and-white photographic sequences captures the movements of 34 different animals as they run, fly, leap, and perform other characteristic actions. Includes 10 bonus Flash animations plus 15 photographic sequences that are ready to be animated.
This book is an animation expert's attempt to ?dump his kit? as the author puts it, and share with others the fundamentals of animation that are not generally taught outside of industry practice, and always were taught behind closed doors.
After photographer Eadweard Muybridge created his revolutionary photographs of animals in motion in the late 1890s, he turned his attention to the study of the human form, by taking detailed photographs in rapid succession step by step as the human body underwent all manner of daily activities. These photographs have served for over a century, as the most highly acclaimed reference point for animators.
The paperback 'lightweight version of 'The Human Figure in Motion', this book as might be extrapolated from the title, contains 60 photographic sequences of the more common everyday activities of both men and women. Most are nude, allowing the musculature and natural drooping from one frame to another to be clearly seen and analysed.
Video games need 'realism boost'
BBC article about how to add authenticity to VR, goes beyond graphics, also encompassing extensive use of motion capture to catalogue how stance, gait and the tiny movements of facial muscles combine when people display different emotions.
Back To Top
Clothing that flows to fit whatever body shape it is on. Fabric that drapes and swirls, covers and exposes, moves in real-time against rigid body models such as avatars and stages. This computationally expensive level of realism is slowly being conquered.
Back To Top
The Uncanny Valley occurs when a face or body looks almost, but not quite real. It is a dip in the increasing levels of empathy you feel towards something that seems more and more human, at the point where it is 'almost there', but just alien enough to throw you off. It does not look right, or does not move right, and this frightens and reviles, on a base level.
Our uncanny ability to spot a fake
A BBC article on the difficulty of maintaining the metaphor inside virtual space, and the issues with uncanny valley that make CGI realism so elusive.
Large Image Display: The Uncanny Valley
 Our look into the original works of Japanese researcher Masahiro Mori and his American counterpart Karl MacDorman, in the mapping out of what we now term the Uncanny Valley, and the nuances within it. Giving hope to the prospect of finally being able to crack the valley and escape it for good.
An expressive face is a work of art. Constantly moving and changing. Lips, brows, frown lines, each is in constant motion. Stop Staring analyses facial structures and movements, then shows animators how to bring life to the faces of their characters.
The depths of the uncanny valley: Dealing with uncanniness
One part of a three-part series on overcoming the Uncanny Valley from a game developer?s perspective; written in 2006. This final part discusses the efforts of others to overcome the valley, and the side effects it has had on the development process.
The Uncanny Valley
 The Uncanny Valley was introduced by Japanese roboticist Masahiro Mori in 1970, as a term referring to the hypothetical valley in which human-like fascimiles suddenly go from 'cute' to 'vile' with almost no warning.
Video games need 'realism boost'
BBC article about how to add authenticity to VR, goes beyond graphics, also encompassing extensive use of motion capture to catalogue how stance, gait and the tiny movements of facial muscles combine when people display different emotions.
Back To TopCrossing the Uncanny Valley (3)
The uncanny valley is a gnarly issue. However, it is not impossible to cross it. Bit by bit, that is just what we are doing; closing the gap between 95% realistic uncanny, and 100% realistic believable.
Large Image Display: Bicentennial Man: Aging Android
 The concept of aging, or the appearance thereof, is a good one. It is an aspect of circumnavigating the uncanny valley that should never be forgotten: No matter how perfectly a human face, behaviour, mannerisms are recreated, unless the face, the body seems to change with time, the uncanny valley has not really been conquered.
The Uncanny Valley was introduced by Japanese roboticist Masahiro Mori in 1970, as a term referring to the hypothetical valley in which human-like fascimiles suddenly go from 'cute' to 'vile' with almost no warning.
Back To TopLiving in the Uncanny Valley (3)
As vital as crossing the valley is to avatar interaction and use, sometimes you can learn as much from an attempt that failed to cross the valley, living miserably within it, than you can from an attempt that managed to fly halfway out. Comparing the failures to the partial successes makes it easier to see what, exactly succeeded, and why.
Large Image Display: The Stepford Wives: Soulless Automation
 A scene from 'The Stepford Wives' in which the robotic skin is shown to Joanna for the first time; along with the knowledge that she will soon be inside it, whether she wishes it or not. This plays on the submersion of humanity within the machine, and also the uncanny valley.
Large Image Display: The Stepford Wives: Stuck in a Loop
 The uncanny valley is a fickle thing. Everything can be fine one moment, and the next, even a slight error in programming, or even a data processing error, triggers it, and you go from believably human to terrifyingly wrong, in a space of only a few seconds.
Back To Top
The humanoid form covers all near-human looking, bipedal forms imaginable. Trolls, ogres, robots, grey aliens and gods. Everything that uses the same basic concepts as the human form, visually.
The Art Models series is three books long and growing. This, the second of the volumes, is smaller than the first, and available in hardback only. Unlike the first book, it shows models of both genders, in a variety of artistic yet practical poses, in full colour and large photos. Contains DVD material.
The Art Models series is three books long and growing. This, the third of the volumes, contains models voted for in a two way process between the authors and the community using the first two books. A wide variety of male and female models in many different poses are present, and appear on the accompanying DVD in 24 different angles per pose.
The Art Models series is three books long and growing. This, the first of the volumes, is the largest. Like the ones that follow it, there are no ultra-skinny models, but a variety of what would be considered average body types, posed nude with no censorship whatsoever.
Not a platform or modelling software specific book, Building a Digital Human is Ken?s second book, and much like the first, it takes you through the basics of creating the 3D forms, in a patient, kind, step by step manner, demonstrating techniques which really aid the beginning to intermediate modeller.
This book is designed for 3-D artists and Web designers who are able to produce interactive 3-D content on their desktop without any prior knowledge of programming. The book uses LifeForms, Virtools and Deep Paint 3-D. It comes with a CD, with most of these tools on it, to aid the reader.
Open Source 3D Human Models
 On the 26th of May 2005, two companies (Zygote Media Group and e frontier) joined forces to give us a new model of open source distribution ? 3D avatars.
Back To Top
Sequence Files in Context (7)
ActiveWorlds: Avatar Movement Restrictions
 As one of the major, established social VR platforms on the internet today, ActiveWorlds deserves more than a cursory look at its world system, its strengths and more importantly, its limitations deserve looking at, in detail. This article focusses on the ActiveWorlds Avatar Control System
Avatar Animation: Why Human(oid)s Swing Their Arms
 Biomedical researchers from the United States and the Netherlands have stated that they can explain why when a human body walks, their arms swing at their sides. To date, this has been duplicated in avatars in virtual environments with hit and miss regularity - half of environments include the feature, half don't.
There always seems to be an artificial disconnect between the skillset necessary to draw in 2D and that necessary to animate with CGI. Many books treat it as if you have to unlearn all you know from one, to learn the other, even when that is clearly not the case. This tome shows why it is not.
One of those rarest of rare books: An animation tome that is not tied down to a specific package. Instead, expect a hefty dose of mathematics, as the author leads you through a variety of animation techniques, both 2D and 3D, from a formulae and expression based perspective.
Swift MoCap Sequence Retrieval
 A better way has been developed for premade MoCap summoning. A method that uses general limb movements to call up the closest matching MoCap file.
Timing for Animation by this pair of authors, is back in print. This book, over 20 years old, was and still is the definitive tome for animation. Whether drawn sculpted or CGI, making something move realistically is all about timing. Timing first, second, and last. When to move, when not to move, what to move, and how much to move.
Back To Top
The creation of realistic looking, realistic moving, flowing, responsive fabrics is one of the holy grails of graphics work. The creation of hair, likewise, is another. Both drape, and flick and flow in ways we are still struggling to recreate.
Back To Top
Rather than starting from scratch each time, there exist programs and libraries designed to give you the basic building blocks. Your creation is still your own, but you can start from a point above scratch. Save time, save effort, save self esteem.
Avatar Pieces for Every VR Environment
 With all graphical elements for 3D VR there is a further issue beyond finding a program or library capable of meeting your requirements. This article describes the simple steps necessary to ensure the items you choose are compatible with your VR of preference.
Back To TopThe Extent of Customisation (1)
When it comes to an avatar, how far is too far when it comes to user customisation? Are there in fact, any limits at all? Should there be?
The Mudder's New Clothes
An interesting, somewhat dated article, on the fascination and burning need participants have in any virtual environment ? textual in this case ? to wear clothing that shows a certain style, even if few if any ever see it. It speaks really, of the tendency in people to add customisation extras to their forms, even if few if any get to see them.
|
