Human Vision and Electronic Imaging 2017
Monday January 30, 2017
Chair Opening Remarks
Session Chairs: Bernice Rogowitz, Visual Perspectives (United States), Thrasyvoulos Pappas, Northwestern University (United States), and Huib de Ridder, Delft University of Technology (the Netherlands)
9:00 – 9:10 AM
Regency Ballroom B
KEYNOTE 1: Vision at a Glance
Session Chair: Bernice Rogowitz, Visual Perspectives (United States)
9:10 – 10:00 AM
Regency Ballroom B
HVEI-111
Vision at a glance (Invited), Ruth Rosenholtz, MIT (United States)
 Ruth Rosenholtz is a Principal Research Scientist in the Dept. of Brain and Cognitive Sciences at MIT. Her lab studies human vision, including visual search, peripheral vision, perceptual organization, and the impact of visual clutter on task performance. Rosenholtz earned her PhD in electrical engineering and computer science, University of California at Berkeley (1994). Prior to MIT, she held research positions with the (Xerox) Palo Alto Research Center, NASA Ames, and with Utrecht University. |
10:00 – 10:30 AM Coffee Break
Special Session: Decoding Visual Semantics: Perceptual Modeling and Deep Learning
Session Chairs:
Jan Koenderink, Katholieke University Leuven (Belgium), and Ruth Rosenholtz, MIT (United States)
10:30 AM – 12:30 PM
Regency Ballroom B
10:30HVEI-112
Eidolons: Effects of capricious local sign (Invited), Jan Koenderink1, Andrea van Doorn2, Matteo Valsecchi3, and Karl Gegenfurtner3; 1Katholieke University Leuven (the Netherlands), 2Utrecht University (the Netherlands), and 3Giessen University (Germany)
11:00HVEI-113
Careful methods and measurements for comparisons between men and machines (Invited), Felix Wichmann1,2, David Janssen1, Robert Geirhos1, Guillermo Aguilar3, Schütt Heiko1, Marianne Maertens3, and Matthias Bethge1; 1Eberhard Karls Universität Tübingen, 2Max-Planck-Institut für Intelligente Systeme, and 3Technische Universität Berlin (Germany)
11:30HVEI-114
Perceptual and engineering implications of cascaded gain control models (Invited), Eero Simoncelli1, Valero Laparra1,2, Johannes Ballé1, and Alexander Berardino1; 1New York University (United States) and 2University of Valencia (Spain)
12:00HVEI-116
Emerging visual representations in deep learning networks (Invited), Aude Oliva, MIT (United States)
12:30 – 2:00 PM Lunch Break
EI 2017 Opening Plenary and Symposium Awards
Session Chairs: Joyce E. Farrell, Stanford University, and Nitin Sampat, Rochester Institute of Technology (United States)
2:00 – 3:00 PM
Grand Peninsula Ballroom D
Giga-scale 3D computational microscopy, Laura Waller, University of California, Berkeley (United States)
Laura Waller is the Ted Van Duzer Endowed Assistant Professor of Electrical Engineering and Computer Sciences (EECS) at UC Berkeley. She is a Senior Fellow at the Berkeley Institute of Data Science, and received her BS (2004), MEng (2005), and PhD (2010) in EECS from the Massachusetts Institute of Technology (MIT). Waller's talk is on computational imaging methods for fast capture of gigapixel-scale 3D intensity and phase images in a commercial microscope that employs illumination-side and detection-side coding of angle (Fourier) space with simple hardware and fast acquisition. The result is high-resolution reconstructions across a large field-of-view, achieving high space-bandwith-time product.
3:00 – 3:30 PM Coffee Break
KEYNOTE 2: Movies and Meaning
Session Chair: Thrasyvoulos Pappas, Northwestern University (United States)
3:30 – 4:20 PM
Regency Ballroom B
HVEI-117
Movies and meaning: From low-level features to mind reading (Invited), Sergio Benini, University of Brescia (Italy)
 Sergio Benini received his MSc in electronic engineering (cum laude) at the University of Brescia (2000) with a thesis granted by Italian Academy of Science. Between ’01 and ’03 he was with Siemens Mobile Communications R&D. He received his PhD in information engineering from the University of Brescia (2006), working on video content analysis. During his PhD he spent one year in British Telecom Research, United Kingdom, working in the “Content & Coding Lab.” Since 2005 he has been an Assistant Professor at the University of Brescia. In 2012, he co-founded Yonder, a spin-off company specialized in NLP, Machine Learning, and Cognitive Computing. |
High Level Vision
Session Chair:
Bernice Rogowitz, Visual Perspectives (United States)
4:20 – 4:40 PM
Regency Ballroom B
4:20HVEI-119
On the role of color in visual saliency, Sergio Etchebehere1 and Elena Fedorovskaya2; 1University Jean Monnet Saint-Etienne (France) and 2Rochester Institute of Technology (United States)
DISCUSSION: From Low-Level Descriptors to Visual Semantics
4:40 – 5:40 PM
Regency Ballroom B
Every afternoon at HVEI, authors from the day’s papers gather to participate in a dynamic discussion with the audience, moderated by the conference and session chairs. Since the papers and the participants represent diverse disciplines, these interactive sessions are exciting and provocative.
5:00 – 6:00 PM All-Conference Welcome Reception, Atrium
Human Vision and Electronic Imaging 2017 Banquet
Hosts:
Bernice Rogowitz, Visual Perspectives (United States), Thrasyvoulos Pappas, Northwestern University (United States), and Huib de Ridder, Delft University of Technology (the Netherlands)
7:15 – 10:00 PM
Sandpebble C,D
Deep learning for gestalt and gestalt for deep learning (Invited), Stella Yu, University of California, Berkeley (United States)
Deep learning has delivered unprecedented success in human vision and computer vision research. The ImageNet classification task which popularized deep learning deals with semantic classification as a simple image-in-label-out supervised learning problem. It essentially asks the same question — "Does this pattern look like what I have seen during training?” — a thousand times. Such a brute-force approach is in a stark contrast to Gestalt psychologists’ view of visual perception — there vision is an emergent property from the inter-relations of visual elements in the scene, but the challenge is what defines visual elements and what are their inter-relations. The distinction between the popular deep learning and the traditional Gestalt approaches boils down to: Do we see a tiger in the scene by going over every object we know of, or by seeing similar (tiger) stripes popping out from the background even if we have no name for what it is? Stella Yu will discuss research that uses deep learning to make Gestalt modeling more practical, and uses Gestalt organization to make deep learning more powerful.
Please join us for an exciting banquet presentation and the opportunity to interact with fellow colleagues and speakers, in an informal atmosphere. You can sign up for the Banquet now, when you register. Gathering and drinks at 7:15 PM, dinner begins at 7:30 PM.
Tuesday January 31, 2017
Human Vision and Stereoscopic Imaging
Session Chairs:
Nicolas Holliman, University of Newcastle (United Kingdom), and Thrasyvoulos Pappas, Northwestern University (United States)
8:50 – 10:10 AM
Grand Peninsula Ballroom D
This session is jointly sponsored by: Stereoscopic Displays and Applications XXVIII and Human Vision and Electronic Imaging 2017.
8:50HVEI-378
Depth-compressed expression for providing natural, visual experiences with integral 3D displays, Yasuhito Sawahata and Toshiya Morita, Japan Broadcasting Corporation (Japan)
9:10HVEI-379
Blind quality prediction of stereoscopic 3D images, Jiheng Wang1, Qingbo Wu2, Abdul Rehman1, Shiqi Wang1, and Zhou Wang1; 1University of Waterloo (Canada) and 2University of Electronic Science and Technology of China (China)
9:30SD&A-380
Pseudo-haptic by stereoscopic images and effects on muscular activity, Takashi Kawai1, Fumiya Ohta1, Sanghyun Kim1, and Hiroyuki Morikawa1,2; 1Waseda University and 2Aoyama Gakuin University (Japan)
9:50SD&A-381
The effects of proximity cues on visual comfort when viewing stereoscopic contents (JIST-first), Yaohua Xie1, Danli Wang2, and Heng Qiao3; 1Chinese Academy of Sciences, 2Institute of Software, Chinese Academy of Sciences, and 3Central University of Finance and Economics (China)
10:00 AM – 7:30 PM Industry Exhibition
10:10 – 10:40 AM Coffee Break
Emerging Issues in Perceptual Image Quality
Session Chair:
Huib de Ridder, Delft University of Technology (the Netherlands)
10:40 AM – 12:40 PM
Regency Ballroom B
10:40HVEI-120
Interactions between saliency and utility, Edward Scott and Sheila Hemami, Northeastern University (United States)
11:00HVEI-121
Perceptual evaluation of psychovisual rate-distortion enhancement in video coding, Zhengfang Duanmu, Kai Zeng, Zhou Wang, and Mahzar Eisapour, University of Waterloo (Canada)
11:20HVEI-122
Balancing Type I errors and statistical power in video quality assessment, Kjell Brunnstrom1,2 and Marcus Barkowsky3; 1Acreo Swedish ICT AB (Sweden), 2Mid Sweden University (Sweden), and 3University of Nantes (France)
11:40HVEI-123
On the perceptual factors underlying the quality of post-compression enhancement of textures, Yusizwan Yaacob, Yi Zhang, and Damon Chandler, Shizuoka University (Japan)
12:00HVEI-124
Do gaze disruptions indicate the perceived quality of non-uniformly coded natural scenes?, Yashas Rai and Patrick Le Callet, University of Nantes (France)
12:20HVEI-125
Subjective evaluation of distortions in first-person videos, Chen Bai and Amy Reibman, Purdue University (United States)
12:40 – 2:00 PM Lunch Break
EI 2017 Tuesday Plenary and Symposium Awards
Session Chairs: Joyce E. Farrell, Stanford University, and Nitin Sampat, Rochester Institute of Technology (United States)
2:00 – 3:00 PM
Grand Peninsula Ballroom D
VR 2.0: Making virtual reality better than reality, Gordon Wetzstein, Stanford University (United States)
Gordon Wetzstein is an Assistant Professor of Electrical Engineering and, by courtesy, of Computer Science, at Stanford University, and leads the Stanford Computational Imaging Group. He received a PhD in computer science from the University of British Columbia (2011) where his doctoral dissertation focused on computational light modulation for image acquisition and display. In his talk, Wetzstein explores the frontiers of VR systems engineering. Eventually, VR/AR systems will redefine communication, entertainment, education, collaborative work, simulation, training, telesurgery, and basic vision research, as next-generation computational near-eye displays evolve to deliver visual experiences that are better than the real world.
3:00 – 3:30 PM Coffee Break
SPECIAL SESSION AND PANEL: Visually Lossless Video Quality for Modern Devices: Research and Industry Perspectives
Panel Moderator: Kjell Brunnstrom, Acreo Swedish ICT AB (Sweden)
Panelists: Damon Chandler, Shizuoka University (Japan); Phil Corriveau, Intel Corporation (United States); Scott Daly, Dolby Laboratories (United States); Edward Delp, Purdue University (United States); and James Goel, Qualcomm Inc. (Canada)
3:30 – 4:20 PM
Regency Ballroom B
3:30HVEI-126
Image and video compression for mobile: Is my screen small enough? (Invited), Edward Delp, Purdue University (United States)
3:40HVEI-127
Business perspectives on perceptually lossless and lossy quality (Invited), Scott Daly, Dolby Laboratories (United States)
3:50HVEI-128
Usage perspectives on perceptually lossless and lossy quality and assessment (Invited), Philip Corriveau1, Juliana Knopf1, Hannah Colett1, and Shun-nan Yang2; 1Intel Corporation and 2Pacific University (United States)
4:00HVEI-129
Subjective assessment and the criteria for visually lossless compression (Invited), Laurie Wilcox1, Robert Allison1, and James Goel2; 1York University and 2Qualcomm Inc. (Canada)
4:10HVEI-130
Masked detection of compression artifacts on laboratory, consumer, and mobile displays (Invited), Yi Zhang, Yusizwan Yaacob, and Damon Chandler, Shizuoka University (Japan)
Moderated Discussion: Visually Lossless Video Quality for Modern Devices: Research and Industry Perspectives
Panel Moderator: Kjell Brunnstrom, Acreo Swedish ICT AB (Sweden)
Panelists: Damon Chandler, Shizuoka University (Japan); Phil Corriveau, Intel Corporation (United States); Scott Daly, Dolby Laboratories (United States); Edward Delp, Purdue University (United States); and James Goel, Qualcomm Inc. (Canada)
4:20 – 5:20 PM
Regency Ballroom B
The issues raised during this panel discussion, and a review of the resulting discussion, will be summarized in a proceedings manuscript, authored by all the contributing panelists:
4:20HVEI-131
Industry and business perspectives on the distinctions between visually lossless and lossy video quality: Mobile and large format displays (Invited), Kjell Brunnstrom1,2, Scott Daly3, Damon Chandler4, Phil Corriveau5, Yi Zhang4, Yusizwan Yaacob4, Laurie Wilcox6, Robert Allison6, James Goel7, Edward Delp8, and Shun-nan Yang9; 1Acreo Swedish ICT AB (Sweden), 2Mid Sweden University (Sweden), 3Dolby Laboratories (United States), 4Shizuoka University (Japan), 5Intel Corporation (United States), 6York University (Canada), 7Qualcomm (Canada), 8Purdue University (United States), and 9Pacific University (United States)
DISCUSSION: Perceptual, Cognitive, and Affective Issues in Image Representation, Compression, and Measurement
5:20 – 6:00 PM
Regency Ballroom B
In this session, authors from the day’s papers will gather to participate in a dynamic discussion with the audience, moderated by the conference and session chairs. Since the papers and the participants represent diverse disciplines, interactive session promised to be exciting and provocative.
5:30 – 7:30 PM Symposium Demonstration Session, Grand Peninsula Ballroom E
Wednesday February 1, 2017
Computational Models of Human Color, Stereo, and High Dynamic Range
8:50 – 10:10 AM
Regency Ballroom B
8:50HVEI-132
Orientation-ocularity maps: A technique for computer vision, Alfredo Restrepo, Universidad de los Andes (Colombia)
9:10HVEI-133
Evaluation of color prediction methods in terms of least dissimilar asymmetric matching, Emitis Roshan and Brian Funt, Simon Fraser University (Canada)
9:30HVEI-134
Characterization of spatiotemporal fluctuation in absorbed light energy by an array of interleaved photosensitive elements, Shahram Peyvandi1, Vebjorn Ekroll2, and Alan Gilchrist1; 1Rutgers, The State University of New Jersey (United States) and 2University of Leuven (KU Leuven) (Belgium)
9:50HVEI-135
Robust dynamic range computation for high dynamic range content, Vedad Hulusic1, Giuseppe Valenzise1, Kurt Debattista2, and Frederic Dufaux1; 1Télécom ParisTech, Université Paris-Saclay (France) and 2University of Warwick (United Kingdom)
10:00 AM – 4:00 PM Industry Exhibition
10:10 – 10:30 AM Coffee Break
Special Session: Computational Modeling Inspired by Human Vision
Session Chair:
Christopher Tyler, Smith-Kettlewell Eye Research Institute (United States)
10:30 AM – 12:30 PM
Regency Ballroom B
10:30HVEI-136
GPU-accelerated vision modeling with the HPE cognitive computing toolkit (Invited), Benjamin Chandler, Hewlett Packard Enterprise (United States)
11:00HVEI-137
A neurally-inspired algorithm for detecting ordinal depth from motion signals in video streams (Invited), Gennady Livitz, Harald Ruda, and Ennio Mingolla, Northeastern University (United States)
11:30HVEI-138
Computational estimation of scene structure through texture gradient cues (Invited), Christopher Tyler1,2 and Ajay Gopi3; 1Smith-Kettlewell Eye Research Institute (United States), 2City University of London (United Kingdom), and 3University of California, Berkeley (United States)
12:00HVEI-139
Learning visual representations for active perception (Invited), Bruno Olshausen, Brian Cheung, and Eric Weiss, University of California, Berkeley (United States)
12:30 – 2:00 PM Lunch Break
EI 2017 Wednesday Plenary and Symposium Awards
Session Chairs: Joyce E. Farrell, Stanford University, and Nitin Sampat, Rochester Institute of Technology (United States)
2:00 – 3:00 PM
Grand Peninsula Ballroom D
Designing VR video camera systems, Brian Cabral, Facebook, Inc. (United States)
Brian Cabral is Director of Engineering at Facebook, leading the Surround 360 VR camera team, specializing in computational photography, computer vision, and computer graphics. He has published a number of papers in the area of computer graphics and imaging including the pioneering Line Integral Convolution algorithm. Cabral discusses developing Facebook Surround 360, an open, high-quality 3D-360 video capture system. VR video capture systems are composed of multiple optical and digital components - all of which must operate as if they are one seamless optical system. The design of VR video cameras, optical choices, SNR, etc., require a new set of technologies and engineering approaches, with tight coupling to the computational system components.
3:00 – 3:30 PM Coffee Break
Image Statistics and Perceptual Features
Session Chairs:
Bernice Rogowitz, Visual Perspectives (United States), and Huib de Ridder, Delft University of Technology (the Netherlands)
3:30 – 4:50 PM
Regency Ballroom B
3:30HVEI-140
Simulation of the biological information available for relative position acuity, Haomiao Jiang1, Nicolas Cottaris2, James Golden1, David Brainard2, Joyce Farrell1, and Brian Wandell1; 1Stanford University and 2University of Pennsylvania (United States)
3:50HVEI-141
Can 'crispening' be explained by contrast gain?, David Kane and Marcelo Bertalmio, Universitat Pompeu Fabra (Spain)
4:10HVEI-142
Defining self-similarity of images using features learned by convolutional neural networks, Anselm Brachmann and Christoph Redies, Jena University Hospital (Germany)
4:30HVEI-143
Determining the influence of image-based cues on human skin gloss perception, Jing Wang1, Jim Mayne2, Carla Kuesten2, Gopa Majmudar2, and Thrasyvoulos Pappas1; 1Northwestern University and 2Amway Corporation (United States)
Digital Humanities: Humans and/vs. Machines
Session Chair:
Thrasyvoulos Pappas, Northwestern University (United States)
4:50 – 5:30 PM
Regency Ballroom B
4:50HVEI-144
Writer identification in modern and historical documents via binary pixel patterns, Kolmogorov-Smirnov test and Fisher's method (JIST-first), Arie Shaus and Eli Turkel, Tel Aviv University (Israel)
5:10HVEI-145
CNN-based transfer learning for historical Chinese seal character recognition based on artificial random samples, Hong Shang, Wei Fan, Jun Sun, and Satoshi Naoi, Fujitsu Research & Development Center (China)
DISCUSSION: Computational Modeling, Perceptual Features, and Digital Humanities
5:30 – 6:20 PM
Regency Ballroom B
In this session, authors from the day’s papers will gather to participate in a dynamic discussion with the audience, moderated by the conference and session chairs. Since the papers and the participants represent diverse disciplines, interactive session promised to be exciting and provocative.
5:30 – 7:00 PM Symposium Interactive Papers (Poster) Session, Atrium
5:30 – 7:00 PM Meet the Future: A Showcase of Student and Young Professionals Research, Atrium
Thursday February 2, 2017
Measuring Fatigue and Discomfort
Session Chair:
Huib de Ridder, Delft University of Technology (the Netherlands)
8:50 – 9:30 AM
Regency Ballroom B
8:50HVEI-146
Comparison of visual discomfort and visual fatigue between HMD and smartphone, Hyeon-Jeong Suk1, Jungmin Han1, and Seon Hee Bae2; 1Korea Advanced Institute of Science and Technology and 2Hansol Eye Clinic (Republic of Korea)
9:10HVEI-147
Measuring visually induced motion sickness using wearable devices, Ran Liu1,2, Eli Peli1, and Alex Hwang1; 1Harvard University (United States) and 2Chongqing University (China)
Attention, Individual Differences, and Emotion
Session Chair:
Bernice Rogowitz, Visual Perspectives (United States)
9:30 – 10:30 AM
Regency Ballroom B
9:30HVEI-148
Developmental changes in ambient and focal visual processing strategies, Onkar Krishna1, Toshihiko Yamasaki1, Kiyoharu Aizawa1, Andrea Helo2, and Pia Rama2; 1The University of Tokyo (Japan) and 2Université Paris Descartes (France)
9:50HVEI-149
Gaze-contingent center-surround fusion of infrared images to facilitate visual search for human targets (JIST-first), Mackenzie Glaholt and Grace Sim, Defence Research and Development Canada (Canada)
10:10HVEI-150
Evaluation and prediction of evoked emotions induced by image manipulations, Lin Yuan and Touradj Ebrahimi, EPFL (Switzerland)
10:30 – 11:00 AM Coffee Break
Special Session: Art and Aesthetics, Part I: Measuring Artistic and Aesthetic Judgments
Session Chairs:
Claus-Christian Carbon, University of Bamberg (Germany), Elena Fedorovskaya, Rochester Institute of Technology (United States), and Monica Lopez-Gonzalez, La Petite Noiseuse Productions (United States)
11:00 AM – 12:30 PM
Regency Ballroom B
11:00HVEI-151
Measurement problems and measurement strategies for capturing the rich experience of art (Invited), Claus-Christian Carbon1,2; 1University of Bamberg and 2EPAEG (Germany)
11:30HVEI-152
The gist of beauty: An investigation of aesthetic perception in rapidly presented images (Invited), Caitlin Mullin1,2, Gregor Hayn-Leichsenring3, Christoph Redies3, and Johan Wagemans2; 1Massachusetts Institute of Technology (United States), 2University of Leuven (Belgium), and 3University of Jena (Germany)
12:00HVEI-153
Gaze patterns in art viewing and their dependency on expertise and image characteristics (Invited), Elena Fedorovskaya, Sanjana Kapisthalam, and Yingtong Bu, Rochester Institute of Technology (United States)
12:30 – 2:00 PM Lunch Break
Special Session: Art and Aesthetics, Part II: Producing Art from the Artists' and Scientists' Perspectives
Session Chairs:
Claus-Christian Carbon, University of Bamberg (Germany), Elena Fedorovskaya, Rochester Institute of Technology (United States), and Monica Lopez-Gonzalez, La Petite Noiseuse Productions (United States)
2:00 – 3:30 PM
Regency Ballroom B
2:00HVEI-154
Imaging human vision: An artistic perspective (Invited), Robert Pepperell, Cardiff Metropolitan University (United Kingdom)
2:30HVEI-155
Art training matters: Enhancement of spatial cognition and brain connectivity (Invited), Lora Likova, Laura Cacciamani, and Spero Nicholas, Smith-Kettlewell Eye Research Instituite (United States)
3:00HVEI-156
Trading conversations between science and art: When musical improvisation enters the dialogue on stage (Invited), Monica Lopez-Gonzalez, La Petite Noiseuse Productions (United States)
3:20 – 4:00 PM Coffee Break
DISCUSSION: Embracing the Complexity of Human Experience
3:30 – 4:30 PM
Regency Ballroom B
In this session, authors from the day’s papers will gather to participate in a dynamic discussion with the audience, moderated by the conference and session chairs. Since the papers and the participants represent diverse disciplines, interactive session promised to be exciting and provocative.
Museum Visit: San Francisco Museum of Modern Art
5:30 – 7:30 PM
Offsite
The San Francisco Museum of Modern Art is open again, after a significant multi-year renovation. To celebrate, we will be organizing a Museum Visit and informal dinner party for Friends of HVEI. Members of the HVEI community will provide perceptual and cognitive insights into pieces in the Museum Collection. Everyone is responsible for his or her own transportation, museum entrance fees and dinner. Anyone interested can meet in the Hotel Lobby at 5:00 to arrange shared cab rides or UBER Ubers. The tour begins at 5:30 at the ticket booth in the museum lobby.
Friends of HVEI Casual Dinner
7:30 – 9:30 PM
Offsite
This casual dinner follows immediately after the museum visit.