现场报道 | SIGGRAPH Asia 2017 (DAY 3):领略前沿Poster papers

今天是SIGGRAPH Asia 2017的第三天,也是Poster papers讲解的最后一天(总共两天,每天中午13:00-14:00)。今年中了poster的文章总共有58篇,分为10个类别,分别有:Animation,Hardware,Image and Video,Interaction,Methods and Applications,Modeling,Multimedia,Rendering,Virtual Environments和Visualization。大部分的文章都附有演示的DEMO和视频,可以帮助快速理解。

这里在文末给大家整理分享论文列表,大家可以自行从网上查找相关资料进行进一步研究。

接下来重点介绍一篇Imaging and Video方向的来自中国北京LlvisionTechnology、NLPR和CASIA团队的文章《Style-Oriented Representative Paintings Selection》。目前根据大师作品的风格来生成或者转化图像的研究和应用比较多。就在昨天的Symposium On Mobile Graphics And Interactive Applications Demonstrations专题上,还展示和交流了一个名为Pictory的作品,App Store有下载,内容就是通过神经网络来实时转换图片风格,通过交流还送了我一张风格化的明信片。

哈哈,跑题了。回到这篇poster文章,它的创意点在于反其道而行之,它的目的是从一堆图像作品中选出最能代表画作作者的作品。它的实现方法是先VGG-19提取图像特征生成content-style feature vector,然后基于自主研发的聚类和抑制框架来对上一步生成的特征向量组进行分类,这里分为两步:1)利用K-means算法基于特征向量组分成K个类;2)然后根据贝叶斯概率分布来找到异类的图片(如下图所示);3)最后,更新这个K个分类并找到具有代表性的图片。

ComputerAnimation Festival

今年ComputerAnimation Festival获奖的作品如下(由于会场不允许拍照,所以无法提供更多的图片和视频):

BESTSTUDENT FILM: GARDEN PARTY

BESTIN SHOW: AFTERWORK

JURYAWARD: HERE’S THE PLAN

JURYAWARD: (OO)

GARDENPARTY讲述的是一只青蛙和一只蟾蜍在废弃的豪宅内自由的探索,最后无意中发现了这个豪宅内最大的秘密。

AFTERWORK设想的是卡通人物在娱乐我们人类的同时,它们自己的生活是怎样的呢?它们是不是也会像我们一样每天下班之后感觉到孤独呢?作者以卡通人物的角度来反思我们现在这种工作和生活的意义。

HERE’STHE PLAN讲述的是一对猫和狗的夫妇梦想开一间自己的蛋糕店,但是由于烤箱坏了它们不得不推迟梦想去挣取更多的钱来购买新的烤箱,但是在某个时刻开始,它们渐渐地偏离了自己的梦想,也偏离了彼此。

(OO)是一个天马行空的作品,正如标题的形状一样,这个作品讲述的是基于两个鼻孔的故事,极大地夸大了我们鼻孔在日常生活中会遇到的各种场景。个人表示欣赏不来这个作品。

今年的Computer Animation Festival还有一个看点是在动画中融入了AR/VR的场景,比如《FLY》这个作品,讲述的是孙子通过VR技术让年迈的爷爷重温了他年轻时开战斗机作战的一些场景(如果没猜错,这场战斗时二战)。

WebAR/WebVR

今天还有一个重头戏是GOOGLE和Mozilla关于WebAR/WebVR的专题报告。先声明一下,这场报告非常多的代码展示和讲解,另外,GOOGLE的工程师还是超有型的,不信你看这个帅哥。

言归正传,他们认为AR/VR的未来其实更像是现在的WEB,可以通过JS接口来与AR/VR设备进行交互,他们开发了three.js/AR.js等并展示了基于WEB的AR/VR应用DEMO。他们目前正在做的是建设更多的原型来更好地探索AR/VR的想象空间。其中包括:具有影响力的示例;制定设计标准;提供沉浸式的接口;渲染技术;研发的工作流程等。让我们拭目以待吧。

附录:Poster Papers

Animation

User Pose Estimation based on multipledepth sensors

SeongminBaek, Myunggyu Kim

Liquid wetting across porous anisotropictextiles

AihuaMao, Mingle Wang, Yong-jin Liu, Huamin Wang, Guiqing Li

4D Computed Tomography Measurement forGrowing Plant Animation

SakihoKato, Tomofumi Narita, Chika Tomiyama, Takashi Ijiri, Hiroya Tanaka

Motion Sickness Simulation Based onSensorimotor Control

Chen-HuiHu, Ping Chen, Yan-Ting Liu, Wen-Chieh Lin

Hardware

Real-time Temporal Quality CompensationTechnique for Head Mounted Displays

Jung-WooChang, Suk-Ju Kang, Min-Woo Seo, Song-Woo Choi, Sang-Lyn Lee, Ho-Chul Lee,Eui-Yeol Oh, Jong-Sang Baek

Multi-DOF 3D Printing with VisualSurveillance

LifangWu, Miao Yu, Yisong Gao, Dong-Ming Yan, Ligang Liu

Aerial Image on Retroreflective Particles

ShinnosukeAndo, Kazuki Otao, Kazuki Takazawa, Yusuke Tanemura, Yoichi Ochiai

Aerial Light-Field Image Augmented BetweenYou and Your Mirrored Image

TomofumiKobori, Kazuki Shimose, Sho Onose, Tomoyuki Okamoto, Masao Nakajima, ToruIwane, Hirotsugu Yamamoto

A Proposal for Wearable Controller Deviceand Finger Gesture Recognition using Surface Electromyography

AyumuTsuboi, Mamoru Hirota, Junki Sato, Masayuki Yokoyama, Masao Yanagisawa

Improvement of A Finger-Mounted HapticDevice Using Surface Contact

Makoto Yoda, Hiroki Imamura

Imagingand Video

Seamless Video Scene Transition UsingHierarchical Graph Cuts

TatsunoriHirai

Style-Oriented Representative PaintingsSelection

YingyingDeng, Fan Tang, Weiming Dong, Hanxing Yao, Bao-Gang Hu

A Deep Convolutional Neural Network forContinuous Zoom with Dual Cameras

Hsueh-IChen, Claire Chen, Yung-Yu Chuang

An Application of a Halftone Pattern Codingin Augmented Reality

BrunoPatrão, Leandro Cruz, Nuno Gonçalves

Subtitle Positioning for E-learning VideosBased on Rough Gaze Estimation and Saliency Detection

BoJiang, Sijiang Liu, Liping He, Weimin Wu, Hongli Chen, Yunfei Shen

Estimating the Simulator Sickness inImmersive Virtual Reality with Optical Flow Analysis

Jiun-YuLee, Ping-Hsuan Han, Ling Tsai, Rih-Ding Peng, Yang-Sheng Chen, Kuan-Wen Chen,Yi-Ping Hung

Interaction

MistFlow : A Fog Display for Visualizationof Adaptive Shape-Changing Flow

KazukiOtao, Takanori Koga

Prototyping Digital Signage Systems withHigh-Low Tech Interfaces

Ting-WenChin, Yu-Yen Chuang, Yu-Ling Fan, Ye-Ning Jiang, Yi-Ching Kang, Wei-Hsin Kuo,Tzu-Wen To, Hiroki Nishino

DupRobo: An Interactive Robotic Platformfor Phyiscal Block-based Autocompletion

TaizhouChen, Yi-Shiun Wu, Feng Han, Baochuan Yue, Kening Zhu

Tactile Braille Learning System to AssistVisual Impaired Users to Learn Taiwanese Braille

Tao-JenYang, Wei-An Chen, Yung-Long Chu, Zi-Xin You, Chien-Hsing Chou

User Interface Applications in Desktop VRusing a Mirror Metaphor

SantawatThanyadit, Ting-Chuen Pong

Deformation Simulation Based on ModelReduction with Rigidity Sampling

Shuo-TingChien, Chen-Hui Hu, Cheng-Yang Huang, Yu-Ting Tsai, Wen-Chieh Lin

Perception Adjustment for Egocentric MovingDistance between Real Space and Virtual Space with See-closed-type HMD

UngyeonYang, Nam-Gyu Kim, Ki-Hong Kim

Dualboard : Integrated User Interfacebetween Typing and Handwriting

SeunghyunWoo, Jia Lee, Changmok Kim

Spin and Roll: Convex Solids of Revolutionas Playful Interface

AkihiroMatsuura, Yuma Ikawa, Yuka Takahashi, Hiroki Tone

Natural Interaction for Media Consumptionin VR Environment

Seung-HwanChoi, Hyun-Jin Kim, Sang-Woong Hwang, Jae-Young Lee

Exploring Mixed-Scale Gesture Interaction

BarrettEns, Aaron Quigley, Hui-Shyong Yeo, Pourang Irani, Thammathip Piumsomboon, MarkBillinghurst

Methodsand Applications

An image generation system of deliciousfood in a manga style

SakikoFujieda, Yuki Morimoto, Kazuo Ohzeki

DeepHolo: Recognizing 3D Objects using aBinary-weighted Computer-Generated Hologram

NaoyaMuramatsu, Chun Wei Ooi, Yuta Itoh, Yoichi Ochiai

Affordable System for MeasuringMotion-to-Photon Latency of Virtual Reality in Mobile Devices

Yu-JuTsai, Yu-Xiang Wang, Ming Ouhyoung

Visualization and Labeling of Point Cloudsin Virtual Reality

JonathanDyssel Stets, Yongbin Sun, Wiley Corning, Scott W. Greenwald

MitsuDomoe: Ecosystem Simulator of VirtualCreatures in Mixed Reality Petri Dish

ToshikazuOhshima, Kenzo Kojima

Haptic Marionette: Wrist Control TechnologyCombined with Electrical Muscle Stimulation and Hanger Reflex

MoseSakashita, Yuta Sato, Ayaka Ebisu, Keisuke Kawahara, Satoshi Hashizume, NaoyaMuramatsu, Yoichi Ochiai

Sketch2Normal: Deep Networks for Normal MapGeneration

WanchaoSu, Xin Yang, Hongbo Fu

Improved and Accessible E-book ReaderApplication for Visually Impaired People

HeesookShin, Youn-Hee Gil, ChoRong Yu, Hee-Kwon Kim, Jisu Lee, Hyung-Keun Jee

Modeling

Polyhedral Meshes with Concave Faces

CaiguiJiang, Renjie Chen

Spring-Pen: reproduction of any softnesswith the 3D printed spring

KengoTanaka, Taisuke Ohshima, Yoichi Ochiai

Bidirectional Pyramid-based PMVS withAutomatic Sky Masking

Jung-JaeYu, Chang-Joon Park

SmartSweep: Context-aware Modeling on aSingle Image

YilanChen, Wenlong Meng, Shi-Qing Xin, Hongbo Fu

Multimedia

A Fast And Efficient Content AwareDownscaling Based Image Compression Method For Mobile Devices

RahulUpadhyay, Ajay Surendranath

Automatic Generation of Visual-TextualWebVideo Thumbnail

BaoquanZha, Shujin Lin, Xin Qi, Zhiquan Zhang, Xiaonan Luo, Ruomei Wang

Aesthetic Temporal and Spatial Editing ofCasual Videos

Jun-HoChoi, Jong-Seok Lee

Content-Based Measure of Image SetDiversity

XingjiaPan, Juntao Ye, Fan Tang, Weiming Dong, Feiyue Huang, Xiaopeng Zhang

Rendering

Spatial Multisampling and MultipassOcclusion Testing for Screen Space Shadows

StefanSeibert, Stefan Radicke

Importance Sampling Measured BRDFs Based onSecond Order Spherical Moment

JieGuo, Yanwen Guo, Jingui Pan

Silk Fabricator: Using Silkworms as 3DPrinters

RikuIwasaki, Yuta Sato, Ippei Suzuki, Atsushi Shinoda, Kenta Yamamot, Kohei Ogawa,Yoichi Ochiai

Eyeshine Rendering: A Real Time RenderingMethod for Realistic Animal Eyes

YunaOmae, Tokiichiro Takahashi

Method for Quantitative Evaluation of theRealism of CG Images Using Deep Learning

MasaakiSato, Masataka Imura

VirtualEnvironments

Emotion Induction in Virtual Environments:A Novel Paradigm Using Immersive Scenarios in Driving Simulators

SeunghyunWoo, Dong-Seon Chang, Daeyun An, Dongjin Hyun, Christian Wallraven, ManfredDangelmaier

Marker-less Real-Time Tracking ofTexture-less 3D objects from a Monocular Image

YukiMorikubo, Naoki Hashimoto

A Spatial User Interface Design UsingAccordion Metaphor for VR Systems

JoungHuemKwon, YoungEun Kim, SangHun Nam

Room VR: A VR Therapy Game For Children WhoFear The Dark

VyDang Ha Thanh, Ondris Pui, Martin Constable

VR Lighting Design

Chi-YangLee, Hsuan-Ming Chang, Chun-Heng Lin, Ming-Han Tsai, Wen-Chieh Lin, Pei-HsienHsu, I-Chen Lin, Yu-Shuen Wang, Jung-Hong Chuang

O-Displaying: An Orientation-BasedAugmented Reality Display on a Smart Glass with a User Tracking from a DepthCamera

Yi-ShanLan, Shih-Wei Sun, Kai-Lung Hua, Wen-Huang Cheng

Immersive VR Environment for ArchitecturalDesign Education

Chia-HungTsou, Ting-Wei Hsu, Chun-Heng Lin, Ming-Han Tsai, Pei-Hsien Hsu, I-Chen Lin,Yu-Shuen Wang, Wen-Chieh Lin, Jung-Hong Chuang

Visualization

Data Jalebi Bot

GauravPatekar, Karan Dudeja

Speech Balloon System Using Single-ChannelMicrophone Array on See-Through Head-Mounted Display

KeiichiZempo, Tomoki Kurahashi, Koichi Mizutani, Naoto Wakatsuki

原文发布于微信公众号 - 腾讯音视频实验室(TencentAVLab)

原文发表时间:2017-12-01

本文参与腾讯云自媒体分享计划,欢迎正在阅读的你也加入,一起分享。

发表于

我来说两句

0 条评论
登录 后参与评论

相关文章

来自专栏CDA数据分析师

原创 | 实战:R环境下Echart的8种可视化

本文由CDA数据分析研究院曾珂提供,刘春娇整理,版权私有,侵权必究,转载请注明出处。 总结一下2016年5月29日数据科学家训练营R语言课程中Echart学习...

23790
来自专栏专知

ACL 2018 计算语言学协会接受论文列表

43410
来自专栏数据科学与人工智能

【陆勤践行】DataSchool 推荐的数据科学资源

Blogs Simply Statistics1: Written by the Biostatistics professors at Johns Hopki...

29890
来自专栏计算机视觉与深度学习基础

计算机视觉著名数据集CV Datasets

Detection PASCAL VOC 2009 datasetClassification/Detection Competitions, Segm...

36480
来自专栏量化投资与机器学习

【最新】新兴市场2018第二季度因子收益报告系列一(附两期下载)

Style Analytics是一家面向投资专业人士的独立全球软件提供商。Style Analytics之前被称为Style Research,拥有超过20年的...

15640
来自专栏HansBug's Lab

1342: [Baltic2007]Sound静音问题

1342: [Baltic2007]Sound静音问题 Time Limit: 5 Sec  Memory Limit: 162 MB Submit: 710 ...

37470
来自专栏目标检测和深度学习

全球最全计算机视觉资料(0:|软件|数据集|挑战赛|创业公司)

47810
来自专栏机器之心

专栏 | 从论文到测试:Facebook Detectron开源项目初探

36560
来自专栏AI科技大本营的专栏

NIPS | 谷歌AI大军来袭,看450多名员工如何横扫今年大会

一年一度的AI盛会NIPS又开始了。 会前数周,就有大神预计,驱车参会的谷歌员工会挤满加州从山景城到长滩的道路,就像这样: ? 图片来源:杜克大学陈怡然教授微博...

45650
来自专栏AI科技大本营的专栏

免费下载 | 20本机器学习和数据科学必读书籍

KDnuggets 网站编辑 Matthew Mayo 特意为广大读者挑选了 20 本机器学习和数据科学相关的书籍。

16030

扫码关注云+社区

领取腾讯云代金券