348 Works

Multi-Scale Point Cloud Analysis

Thibault Lejemble, Claudio Mura, Loïc Barthe & Nicolas Mellado
Surfaces sampled with point clouds often exhibit multi-scale properties due to the high variation between their feature size. Traditional shape analysis techniques usually rely on geometric descriptors able to characterize a point and its close neighborhood at multiple scale using smoothing kernels of varying radii. We propose to add a spatial regularization to these point-wise descriptors in two different ways. The first groups similar points in regions that are structured in a hierarchical graph. The...

Adaptive Frameless Rendering with NVIDIA OptiX

Chung-Che Hsiao & Benjamin Watson
We implement adaptive frameless rendering (AFR) on NVIDIA OptiX, a real-time ray tracing API taking advantage of NVIDIA GPUs including their latest RTX functionality. OptiX is a parallel system that sits on top of NVIDIA's better-known CUDA API. AFR has sampling and reconstruction processes that use information distributed across both space and time, aiming to generate low-latency updates. Previous AFR implementations were sequential prototypes. Our parallel prototype is allowing us to confront several unique challenges,...

Can we Invert a Local Reflectance Model From a Single Specular Highlight with Known Scene Geometry and Camera Pose?

Souheil Hadj-Said, Mohamed Tamaazousti & Adrien Bartoli
Recovering the scene's illumination from images is a crucial step in Augmented and Diminished Reality. We present an experimental investigation to assess the well-posedness of this problem (i) using a known scene geometry and camera pose and (ii) assuming that a local reflectance model such as Blinn-Phong's holds. Based on results on synthetic and real data, we establish two major observations. First, the problem of retrieving the full local reflectance model's parameters and light source...

Towards the Fitting of Parametric 2D Sketches and 3D CAD Models to Point Clouds of Digitized Assemblies for Reverse Engineering

Ghazanfar Ali Shah, Franca Giannini, Marina Monti, Arnaud Polette & Jean-Philippe Pernot
Following commonly used reverse engineering techniques, it is very difficult to reconstruct editable CAD parts or assemblies that can later be used and modified in the Product Development Process (PDP). Traditional methods follow a sequential time-consuming patch-by-patch reconstruction strategy with cumbersome procedures in which designers usually have to face many issues (e.g. decomposition in patches, trimming and connection of the patches), and generally producing ''dead'' models that cannot be later modified as needed. This paper...

VITON-GAN: Virtual Try-on Image Generator Trained with Adversarial Loss

Shion Honda
Generating a virtual try-on image from in-shop clothing images and a model person's snapshot is a challenging task because the human body and clothes have high flexibility in their shapes. In this paper, we develop a Virtual Try-on Generative Adversarial Network (VITON-GAN), that generates virtual try-on images using images of in-shop clothing and a model person. This method enhances the quality of the generated image when occlusion is present in a model person's image (e.g.,...

Voxelizing Light-Field Recordings

David C. Schedl, Indrajit Kurmi & Oliver Bimber
Light fields are an emerging image-based technique that support free viewpoint navigation of recorded scenes as demanded in several recent applications (e.g., Virtual Reality). Pure image-based representations, however quickly become inefficient, as a large number of images are required to be captured, stored, and processed. Geometric scene representations require less storage and are more efficient to render. Geometry reconstruction, however, is unreliable and might fail for complex scene parts. Furthermore, view-dependent effects that are preserved...

Learning Generative Models of 3D Structures

Siddhartha Chaudhuri, Daniel Ritchie, Kai Xu & Zhang, Hao (Richard)
Many important applications demand 3D content, yet 3D modeling is a notoriously difficult and inaccessible activity. This tutorial provides a crash course in one of the most promising approaches for democratizing 3D modeling: learning generative models of 3D structures. Such generative models typically describe a statistical distribution over a space of possible 3D shapes or 3D scenes, as well as a procedure for sampling new shapes or scenes from the distribution. To be useful by...

libigl: Prototyping Geometry Processing Research in C++

Daniele Panozzo & Alec Jacobson
Modern geometry processing algorithms depend on an ever-growing toolbox of fundamental sub-routines and data structures. Prototyping from scratch requires much time building basic tools rather than focusing on the novel research idea. Many existing code libraries have unsatisfactory APIs and the time spent implementing sub-routines is often replaced with time spent learning complex, templated object hierarchies or memory layouts. Libigl is a C++ library of geometry processing algorithms designed for and by researchers. Its wide...

Deep Learning for Computer Graphics and Geometry Processing

Michael Bronstein, Leonidas Guibas, Iasonas Kokkinos, Or Litany, Niloy Mitra, Federico Monti & Emanuele Rodolà
In computer graphics and geometry processing, many traditional problems are now becoming increasingly handled by data-driven methods. In an increasing variety of problem settings, deep networks are state-of-the-art, beating dedicated hand-crafted methods by significant margins. This tutorial gives an organized overview of core theory, practice, and graphics-related applications of deep learning.

Smoothed Particle Hydrodynamics Techniques for the Physics Based Simulation of Fluids and Solids

Dan Koschier, Jan Bender, Barbara Solenthaler & Matthias Teschner
Graphics research on Smoothed Particle Hydrodynamics (SPH) has produced fantastic visual results that are unique across the board of research communities concerned with SPH simulations. Generally, the SPH formalism serves as a spatial discretization technique, commonly used for the numerical simulation of continuum mechanical problems such as the simulation of fluids, highly viscous materials, and deformable solids. Recent advances in the field have made it possible to efficiently simulate massive scenes with highly complex boundary...

A Breadth-First Introduction to VFX: A Holistic Approach for Teaching the Visual Effects Production Pipeline

Adam Redford, Melania Fodritto & Eike Falk Anderson
This paper outlines a tried and tested breadth-first approach for teaching essential key areas of the visual effects post production pipeline to novices. Successfully used in a first semester course our holistic process guides the creation of a single self-contained visual effects shot (video sequence) of photorealistic quality. We provide an overview of course curriculum, organization and exercises (lessons covering the essential areas of visual effects production for creating each of the individual elements of...

Vegetable Matter Decay: An Exceptional Student 'Innovations' Project

Eike Falk Anderson, Bianca Cirdei & Valery Adzhiev
We present an outstanding undergraduate research project that has resulted in the development of a parametrised method for simulating drying and decaying vegetable matter from the fruits category, taking into account the biological characteristics of the decaying fruit. This work was awarded 1st place in the undergraduate category of the ACM Student Research Competition at SIGGRAPH 2018.

Enrichment of TV Series With Augmented and Mixed Reality Applications on Mobile Devices and on HoloLens

R. Strzebkowski, T. Gehrmann & D. Wulf
In this demo we will present two related AR/MR projects realized in cooperation with the German public broadcaster ZDF, which consist of an interactive mobile 3D Augmented Reality infotainment/game application as well as a Mixed Reality version of the infotainment app for HoloLens. The aim of the projects was to evaluate the technical, gamification and infotainment potentials for so-called companion apps based on AR/MR technologies for the very successful TV series 'Die Bergretter' (The Mountain...

Virtual Agent System to Induce Reminiscence Utterances With the Past and Present Map Information

Yutaka Takase, Yukiko I. Nakano & Kiyoshi Yasuda
In recent decades, the increasing number of elderly people with dementia has become a major social problem. Under such circumstances, virtual agent systems aimed at helping with mental health care for the elderly have attracted increasing attention. In this study, we propose a novel virtual agent system that encourages users' speaking about their reminiscences of their hometown or favorite places. The key idea is that the contents that the agent delivers (i.e., speaks) are generated...

Natural Interface and Overreaction for VR Entertainment

T. Kawamoto, T. Horikoshi & N. Liberati
With the widespread use of inexpensive VR devices, many VR events for entertainment started to be organized all over the world. An important point when many people experience VR content is easily engage the experience without losing too much time to understand how the interaction between users and virtual world is designed. Thus, it is very important to think of simple and intuitive interfaces. We aimed for a more intuitive interaction, and we replaced the...

VirtualCar: Virtual Mirroring of IoT-Enabled Avacars in AR, VR and Desktop Applications

G. Pappas, J. Siegel & K. Politopoulos
This demo features a virtual 3D car model based off a vehicle's data duplicate, an IoT-enabled Avacar. There are three embodiments including a desktop application (VirtualCar), Augmented Reality (VirtualCarAR), and Virtual Reality (VirtualCarVR). The Avacar gets information from an external API serving information from a real or rep-resentative simulated vehicle, mirroring the sensors present in modern On-Board Diagnostic systems and the loca-tion data provided by a mobile device's GPS. In each application, the Avacar appears...

Study of Human Behavioral Characteristics: Using Virtual Reality to Analyze Behavior of Following Moving Objects

Kazuya Yonezawa, Eigo Segawa & Ryuta Tanaka
It is crucial to model the physical world for simulation in which various phenomena are represented on a computer. In the case of simulation involving human actions, it is necessary to model not only physical phenomena but also human behaviors. However, there are still many unclear things about human behaviors. In this study, to clarify these behavioral characteristics, we verified two hypotheses using virtual reality. One was that ''people follow moving objects similar to humans...

Visuo-Haptic Redirected Walking Using Handrail

Keigo Matsumoto, Takuji Narumi, Yohei Yanase, Yuki Ban, Tomohiro Tanikawa & Michitaka Hirose
A novel use of virtual reality (VR) was demonstrated. It enables a user to feel walking a straight virtual corridor while touching straight handrails, although the user walks while gripping circular handrails within a limited space. The main contribution of this study is to show that visual-haptic redirect walking is possible using handrails for a haptic presentation. Matsumoto et al. introduced visuo-haptic redirected walking. They used walls as a haptic presentation, and they produced Unlimited...

User Evaluation of Visuo-Haptic Collocation System

B. Cha, Y. Bae, W. Choi & J. Ryu
In the encounter-type haptic display systems, visuo-haptic collocation is essential because a poorly collocated pair of virtual and real objects may hinder natural interaction. This poster presents user evaluation results of a visuo-haptic collocation system. By measuring performances of two tasks for pressing a button and grasping a steering wheel, the developed visuohaptic collocation system can provide users with natural interaction experiences.

Effect of an Encounter-Type Haptic Device on Forceps Position Reproducibility in Laparoscopic Tracking Training

Akinobu Morishima, Akihiro Terashima, Kazutaka Obama, Yoshiharu Sakai, Taro Maeda & Hideyuki Ando
This paper describes an improvement to our previously proposed laparoscopic training system that was designed to improve surgical skill transmission by allowing surgical trainees to replicate the movements of skilled surgeons dis-played on a monitor. However, we found that with this system, the precision of followability in the depth direction was low in comparison with that on the monitor in the planar direction. Herein, to improve the reproducibility of the forceps position and present clues...

Contribution to Ownership and Agency When Controlling a Mobile Robot

Junjie Hua, Masahiro Furukawa & Taro Maeda
Out-of-body experience shows that body ownership can be built through illusory visual-tactile correlation when man observes avatar from its back.On the other hand, ownership and agency are usually thought to be negative when observing avatar face to face. We aim to clarify whether person perspective plays an important role on ownership and agency when controlling a mobile robot, and if so, how does ownership and agency change when person perspective changes. In this article, we...

Registration Year

  • 2019

Resource Types

  • Text