CPSC 526 Class Wiki

Test post for paper discussion.

-- MichielVanDePanne - 04 Nov 2011

Contribution:

Virtual characters are represented as directed graphs having nodes and connection. A genetic language based evolution of motor as well as neural control is suggested. No user information regarding shape/size, joint constraints etc are assumed when developing the creatures.

Computation of results:

Since the entire task is computationally expensive, parallel implementation of the task is done. Evolutions of about 50 to 100 generations i.e. by mating directed graphs either by crossover operation (30%) or by grafting(30%) and sometimes asexually (only using mutations) (40%), are performed corresponding to four types of behavior, i.e. swimming, jumping, walking and following. Another suggested method was to interactively evolve a morphology so as the creatures have aesthetic touch to them instead of just goal based evolution. Hybrid world creatures were also developed.

Reproducibility:

My view is that the paper is reproducible. Although models are not detailed mathematically, they make up for it by giving exhaustive description of the same. The models used for sensors, nerons and Effectors are detailed properly in section 3. Details about the use of physical simulation techniques e.g Feather's method to calculate accelerations, Runge-Kutta fehlberg method for integration and other details for virtual world simulation (e.g water, collison etc) are provided in section 4. The genetic optimization based behavior selection method is also well explained in section 5.

Improvements in Research/Writing:

Overall the paper is well written but I would have preferred some more light to be thrown on the technical aspects of the overall model and specially the modeling of characters as directed graphs and the genetic evolution process.

-- Main.sumanm - 17 Nov 2011

Above is the review of : "Evolving Virtual Creatures"

-- Main.sumanm - 17 Nov 2011

Review : "Practical Character Physics for Animators" Contribution: - Interactive system to enhance character motions, by incorporating physical properties into a existing key frame system. - Tools to enhance realism of balistic motions. - Investigate physical accuracy of manually created animation by skilled animators.

Result Evaluation:

1. Ballistic Paths : The path is created based upon the center of mass m and a constraint. The animator has 2 ways to incorporate this into the animation, either by adjusting the timing of the motion, freezing the initial and final positions or manualy adjust the path and then retime the entire motion.

2. Angular Momentum: An iterative approximation algorithm is suggested to compute the configuration (T_o) of the character given the momentum (h) specified by the user. THe method makes use of the rigid body inertia (J), velocity (v_i), the position (x)

Reproducibility:

The paper should be reproducible to a considerable extent. All the techniques used are backed up with corresponding pseudo code/step-by-step algorithm. The variables are clearly explained in both the cases of ballistic path and angular momentum. An appendix explaining the background on the equations governing the physical properties of an animation is also provided for assistance.

-- Main.sumanm - 17 Nov 2011

"Practical Character Physics for Animators"

a) Contribution This paper contributes an extension of the traditional keyframe-based animation authoring interface that includes visualizations of physically correct motions. The goal is to provide tools and feedback to animators to allow them to produce more physically accurate motions, without removing any of the animator's control.

b) Evaluation The system is evaluated by having professional animators use the tool, and by comparing the physical accuracy of animations produced with and without use of the tool.

c) Reproducibility The tools described in the paper are certainly reproducible, as the algorithms used are given in sufficient detail to recreate them. The success of the tool in terms of usefulness to professional animators is not as certain to be reproducible, since not many details of the authors' specific evaluation methodologies are explained in the paper.

d) Improvements There are a number of spelling and grammar errors present in the paper that require correction in order to improve readability and understandability. Furthermore, the figures and tables are all clumped at the end rather than located near the relevant section of the paper's body, which also makes it more difficult to read and understand exactly what the research is accomplishing.

-- Main.cdoran - 17 Nov 2011

Evolving Virtual Creatures

What is the contribution of the paper?

In previous work, control system must be designed for each type of fixed structure. The method in this paper can generate physical structure and its corresponding control system automatically in a similar fashion to creature evolution. So the generated physical structure and the control system is also called a creature. Users don't have to specify any parameters of physical structure or control system. The generated physical structure and control system work in 3D physics.

How are the results evaluated?

The results are evaluated by looking at the generated creatures in various environments. An important concern in the evaluation seem to be how well creatures can adapt to the environment by performing expected tasks, i.e. does it swim or walk fast, or does it jump high?

It is also evaluated by the number of generated creatures and the time it takes to do so.

Is the paper reproducible?

Somewhat reproducible, but it's not clear to what genotypes look like and how they can be converted to creature morphology and creature control.

how could the paper research or paper writing be improved?

Overall, this paper is well written, but in the introduction, there is not an overview of sections two through six and how those sections are related to one another. It would be nice if they do so.

Practical Character Physics for Animators

What is the contribution of the paper?

Traditionally, animators use keyframing and inverse kinematics to animate characters. This paper develops an interactive system that allows animators to create physics-based characters which makes the motion more physically plausible.

How are the results evaluated?

The results are evaluated by the physical realism of the animation.

Is the paper reproducible?

It looks to be reproducible, given the equation to compute ballistic paths, and those equations and the algorithm needed to develop the angular-momentum tool.

how could the paper research or paper writing be improved?

Section subsection headers can be more meaningful. For example, section two "Improving Physical Realism" and section three "Keyframe Animations' Physical Accuracy" look similar to each other and they don't seem to summarize the respective sections very well.

Main.shuoshen - 17 Nov 2011

--Practical Character Physics For Animators

• contribution: It was the first research that incorporated physical properties into an existing key frame-based system by the time the research was conducted. Tools were developed to help animators modify ballistic motions of characters and the physical accuracy of the high quality animations was investigated by measuring important physical properties.

• evaluation: Basically, the tools were tested by animators and the evaluation was based on the effectiveness of how much it helped the animators to create the physical realistic animation.

• reproducibility: I think the tool is reproducible since the basic idea is not hard to understand, and the math details are provided along with appendix. However some of the results seem to be less reproducible since it requires professional animators to create the ballistic motions, which could be uncertain.

• comments: Overall the paper is not hard to follow. However the structure of the paper is not clear when I first read it. I have to go back and forth to get clear about the organization. I also encountered a few issues. For example, the term “COM” was not explained, at the beginning I could only guess it is “center of mass” which could also be “center of momentum”. The paper also has a number of grammar or spelling issues such as “derived form”. But given the factor that this is a preprint version, the author may have corrected these kinds of issues in the final version.

--Evolving Virtual Creature

• contribution: The research creates a novel system for creating virtual creatures behaving in 3D physical world. The creature’s morphologies as well as their control systems are evolved. “A genetic language is presented that uses nodes and connections as its primitive elements to represent directed graphs, which are used to describe both the morphology and the neural circuitry of these creature.”[Karl Sims]

• evaluation: The experiment runs in parallel with one master node performing the genetic algorithm that chosen randomly from asexual, crossovers and grafting, and then sending out genotypes to other slave nodes to do fitness test and then gathering back the fitness values. The fitness tests include swimming, walking, jumping and following. After 50 to 100 generations, creatures that pass the fitness threshold survive and get selected.

• reproducibility: In my opinion the paper should be reproducible however it will involve large amount of work since the models used in the research are presented without detailed mathematics explanation. And I still don’t quite get how the phenotype “brain” is generated. But I think the physical simulation part shouldn’t be that hard since we use simplified models.

• comments: Overall the paper is well organized and I found the structure is not hard to follow. However it takes time to learn some background knowledge.

Baoxuan Xu

Practical Character Physics For Animators (Ben Humberston comments)

Contribution

The authors present a group of tools that provides visual feedback about the physical accuracy of an animation during character animation authoring. They provide three specific tools:

• Ballistic trajectory: A simple visualization of a point-particle trajectory through the air; the user may modify the start point, end point, time in air, and initial velocity to achieve a desired path, then modify their existing keyframed animation to better fit the ballistic path

• Angular Momentum: The character's angular momentum during free flight is calculated, and the character's root orientation may be adjusted to preserve the calculated value.

• COP/COM: The projection of a character's 3D center of mass/pressure is shown on a flat 2D plane beneath the character along with the support polygon of the feet.

Additionally, the authors provide observations on the physical errors in animations created by professionals without the use of the paper's tools.

Evaluation

The authors only provide subjective evaluation of their system's usefulness for animators. They put strong emphasis on it's effectiveness as a standardizing tool among multiple animators in a studio, suggesting that the satisfaction experienced by solo animators when using the tool may be quite variable. Additionally, they provide no controlled evaluations of perceived animation quality that compares animations produced with and without their tools.

On the other hand, they do provide quantitative observations of physical errors made by animators without the use of the author's tools. These observations help separate instances where their tool is likely superfluous from those where it could significantly improve the physical accuracy of animations

Reproducibility

The authors provide extensive detail on the mathematical basis of their ballistic path and angular momentum tools. Though the algorithms are relatively simple, they do provide pseudocode in figures. However, because the paper concerns tools provided to human users, the ease-of-use and clarity of the interface shown to the user by the tools is perhaps more important than the technical details; the complete UI system may be difficult to reproduce without more details.

Improvements

• The evaluation of the tools' effectiveness for animators is fairly vague; quantitative results or more controlled testing may have yielded more definite results.

• The angular momentum tool is only capable of modifying the root orientation of the character to correct the value; a more sophisticated system with the option of modifying character joint angles as well may have been useful when a character's rotation speed is fixed

-- BenHumberston - 17 Nov 2011

-- Main.davidm - 18 Nov 2011

Evolving Virtual Creatures - David Matheson

(a) Contribution: The key contribution is a framework for applying genetic algorithms to optimize creature morphologies and controller systems. They represent both with directed graphs where the nodes specify rigid bodies and partial "neural nets" (they're not really neural nets).

(b) Evaluation: The results were evaluated qualitatively. For each task they describe different morphologies and control strategies that were selected. It would have been nice to see the relative performance of different morphology and control schemes.

(c) Reproducibility: The idea of the paper is reproducible and their evaluation functions are well documented. However the parameters for the possible rigid bodies are not specified. Also the parameters for the simulation such as high and low velocities for switching between spring/impulse models are not provided.

(d) Improvements in Research/Writing: This paper is a framework for applying genetic algorithms to morphology and control systems. Different meta control schemes could be evaluated using different models for control. As long as it maintains the requirements of the genetic selection it will still work as part of the framework. I think the paper could have been clearer by specifying the requirements for applying genetic selection to morphology and control independently of their particular specification of morphology and control.

Practical Character Physics for Animators - David Matheson

(a) Contribution: An interactive system that allows animators to combine their traditional kinematic approach with physical simulation. They focus on visualizing realistic ballistic trajectories and angular momentum. This allows animators to use the algorithms specified to adjust their animations to make them more physically correct.

(b) Evaluation: The results outline how physically realistic a set of animations are before and after using the tools specified in the paper. While they provide quantitative results they don't provide the number of animators/animations evaluated. This paper is evaluating the animators as much as the their purposed tools.

(c) Reproducibility: Yes the tools are reproducible but their claims of how animators use them are not as reproducible.

(d) Improvements in Research/Writing: Overal the paper is well written but it is more like a technical report than a research paper. They go through the math in good detail. In some cases other papers are referenced to explain terminology (such preparatory and recovery phases) when it could have been explained as part of this paper.