38-Issue 1

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2632737

Browse

Browsing 38-Issue 1 by Subject "animation"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Flexible Use of Temporal and Spatial Reasoning for Fast and Scalable CPU Broad‐Phase Collision Detection Using KD‐Trees
    (© 2019 The Eurographics Association and John Wiley & Sons Ltd., 2019) Serpa, Ygor Rebouças; Rodrigues, Maria Andréia Formico; Chen, Min and Benes, Bedrich
    Realistic computer simulations of physical elements such as rigid and deformable bodies, particles and fractures are commonplace in the modern world. In these simulations, the broad‐phase collision detection plays an important role in ensuring that simulations can scale with the number of objects. In these applications, several degrees of motion coherency, distinct spatial distributions and different types of objects exist; however, few attempts have been made at a generally applicable solution for their broad phase. In this regard, this work presents a novel broad‐phase collision detection algorithm based upon a hybrid SIMD optimized KD‐Tree and sweep‐and‐prune, aimed at general applicability. Our solution is optimized for several objects distributions, degrees of motion coherence and varying object sizes. These features are made possible by an efficient and idempotent two‐step tree optimization algorithm and by selectively enabling coherency optimizations. We have tested our solution under varying scenario setups and compared it to other solutions available in the literature and industry, up to a million simulated objects. The results show that our solution is competitive, with average performance values two to three times better than those achieved by other state‐of‐the‐art AABB‐based CPU solutions.Realistic computer simulations of physical elements such as rigid and deformable bodies, particles and fractures are commonplace in the modern world. In these simulations, the broad‐phase collision detection plays an important role in ensuring that simulations can scale with the number of objects. In these applications, several degrees of motion coherency, distinct spatial distributions and different types of objects exist; however, few attempts have been made at a generally applicable solution for their broad phase. In this regard, this work presents a novel broad‐phase collision detection algorithm based upon a hybrid SIMD optimized KD‐Tree and sweep‐and‐prune, aimed at general applicability. Our solution is optimized for several objects distributions, degrees of motion coherence and varying object sizes. These features are made possible by an efficient and idempotent two‐step tree optimization algorithm and by selectively enabling coherency optimizations.
  • Thumbnail Image
    Item
    A Probabilistic Steering Parameter Model for Deterministic Motion Planning Algorithms
    (© 2019 The Eurographics Association and John Wiley & Sons Ltd., 2019) Agethen, Philipp; Gaisbauer, Felix; Rukzio, Enrico; Chen, Min and Benes, Bedrich
    The simulation of two‐dimensional human locomotion in a bird's eye perspective is a key technology for various domains to realistically predict walk paths. The generated trajectories, however, are frequently deviating from reality due to the usage of simplifying assumptions. For instance, common deterministic motion planning algorithms predominantly utilize a set of static steering parameters (e.g. maximum acceleration or velocity of the agent) to simulate the walking behaviour of a person. This procedure neglects important influence factors, which have a significant impact on the spatio‐temporal characteristics of the finally resulting motion—such as the operator's physical conditions or the probabilistic nature of the human locomotor system. In overcome this drawback, this paper presents an approach to derive probabilistic motion models from a database of captured human motions. Although being initially designed for industrial purposes, this method can be applied to a wide range of use cases while considering an arbitrary number of dependencies (input) and steering parameters (output). To underline its applicability, a probabilistic steering parameter model is implemented, which models velocity, angular velocity and acceleration as a function of the travel distances, path curvature and height of a respective person. Finally, the technical performance and advantages of this model are demonstrated within an evaluation.The simulation of two‐dimensional human locomotion in a bird's eye perspective is a key technology for various domains to realistically predict walk paths. The generated trajectories, however, are frequently deviating from reality due to the usage of simplifying assumptions. For instance, common deterministic motion planning algorithms predominantly utilize a set of static steering parameters (e.g. maximum acceleration or velocity of the agent) to simulate the walking behaviour of a person. This procedure neglects important influence factors, which have a significant impact on the spatio‐temporal characteristics of the finally resulting motion—such as the operator's physical conditions or the probabilistic nature of the human locomotor system. In overcome this drawback, this paper presents an approach to derive probabilistic motion models from a database of captured human motions. Although being initially designed for industrial purposes, this method can be applied to a wide range of use cases while considering an arbitrary number of dependencies (input) and steering parameters (output).
  • Thumbnail Image
    Item
    Style Invariant Locomotion Classification for Character Control
    (© 2019 The Eurographics Association and John Wiley & Sons Ltd., 2019) Boehs, G.E.; Vieira, M.L.H.; Chen, Min and Benes, Bedrich
    We present a real‐time system for character control that relies on the classification of locomotive actions in skeletal motion capture data. Our method is both progress dependent and style invariant. Two deep neural networks are used to correlate body shape and implicit dynamics to locomotive types and their respective progress. In comparison to related work, our approach does not require a setup step and enables the user to act in a natural, unconstrained manner. Also, our method displays better performance than the related work in scenarios where the actor performs sharp changes in direction and highly stylized motions while maintaining at least as good performance in other scenarios. Our motivation is to enable character control of non‐bipedal characters in virtual production and live immersive experiences, where mannerisms in the actor's performance may be an issue for previous methods.We present a real‐time system for character control that relies on the classification of locomotive actions in skeletal motion capture data. Our method is both progress dependent and style invariant. Two deep neural networks are used to correlate body shape and implicit dynamics to locomotive types and their respective progress. In comparison to related work, our approach does not require a setup step and enables the user to act in a natural, unconstrained manner. Also, our method displays better performance than the related work in scenarios where the actor performs sharp changes in direction and highly stylized motions while maintaining at least as good performance in other scenarios. Our motivation is to enable character control of non‐bipedal characters in virtual production and live immersive experiences, where mannerisms in the actor's performance may be an issue for previous methods.
  • Thumbnail Image
    Item
    Urban Walkability Design Using Virtual Population Simulation
    (© 2019 The Eurographics Association and John Wiley & Sons Ltd., 2019) Mathew, C. D. Tharindu; Knob, Paulo R.; Musse, Soraia Raupp; Aliaga, Daniel G.; Chen, Min and Benes, Bedrich
    We present a system to generate a procedural environment that produces a desired crowd behaviour. Instead of altering the behavioural parameters of the crowd itself, we automatically alter the environment to yield such desired crowd behaviour. This novel inverse approach is useful both to crowd simulation in virtual environments and to urban crowd planning applications. Our approach tightly integrates and extends a space discretization crowd simulator with inverse procedural modelling. We extend crowd simulation by goal exploration (i.e. agents are initially unaware of the goal locations), variable‐appealing sign usage and several acceleration schemes. We use Markov chain Monte Carlo to quickly explore the solution space and yield interactive design. We have applied our method to a variety of virtual and real‐world locations, yielding one order of magnitude faster crowd simulation performance over related methods and several fold improvement of crowd indicators.