Program
Keynote Speakers
Conference Program
3DV 2018 will showcase high quality single-track oral and poster presentations, and demonstration sessions. The main conference will run from 5 to 7 September, in conjunction with the industrial exhibition, followed by the tutorials on 8 September.
Conference booklet available here.
Keynote Session 1
Wed. 5/9 10:00 - 10:45. Session Chair: Tomas Pajdla
-
Motion Averaging in 3D Vision
Oral Session 1.A
Wed. 5/9 11:15 - 12:00. Session Chair: Guy Godin
-
Keep it Unreal: Bridging the Realism Gap for 2.5D Recognition with Geometry Priors Only
-
Surface Light Field Fusion
-
Joint Material and Illumination Estimation from Photo Sets in the Wild
Spotlight Session 1
Wed. 5/9 12:00 - 13:15. Session Chair: Massimiliano Corsini
-
Fusion++: Volumetric Object-Level SLAM
-
Patch-based Non-rigid 3D Reconstruction from a Single Depth Stream
-
Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation
-
Multi-scale CNN stereo and pattern removal technique for underwater active stereo system
-
A Multi-Resolution Approach for Color Correction of Textured Meshes
-
Progressive Large-Scale Structure-from-Motion with Orthogonal MSTs
-
Multi-Scale Direct Sparse Visual Odometry for Large-Scale Natural Environment
-
Detailed Human Avatars from Monocular Video
-
DeepHPS: End-to-end Estimation of 3D Hand Pose and Shape by Learning from Synthetic Depth
-
Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB
-
Learning to Align Images using Weak Geometric Supervision
-
Structure-Aware Shape Synthesis
-
Learning Material-Aware Local Descriptors for 3D Shapes
-
Joint Representation of Primitive and Non-Primitive Objects for 3D Vision
-
Joint Alignment of Multiple Generalized Point Sets With Anisotropic Positional Uncertainty Based on Expectation Maximization
-
(IGSP) Iterative Global Similarity Points : A robust coarse-to-fine integration solution for pairwise 3D point cloud registration
-
SegmentedFusion: 3D human body reconstruction using stitched bounding boxes
-
Learning to Estimate Indoor Lighting from 3D Objects
Poster Session 1
Wed. 5/9 14:45 - 16:30.
All the papers presented on 5/9
Oral Session 1.P
Wed. 5/9 16:30 - 17:45. Session Chair: Frank Dellaert
-
Camera Orientation Estimation in Natural Scenes Using Semantic Cues
-
Four- and Seven-Point Relative Camera Pose from Oriented Features
-
Plane-Based Optimization of Geometry and Texture for RGB-D Reconstruction of Indoor Scenes
-
RecResNet: A Recurrent Residual CNN Architecture for Disparity Map Enhancement
-
MVDepthNet: Real-time Multiview Depth Estimation Neural Network
Keynote Session 2
Thu. 6/9 9:00 - 9:45. Session Chair: Andrea Torsello
-
Hololens [tentative title]
Oral Session 2.A
Thu. 6/9 10:15 - 11:00. Session Chair: Akihiro Sugimoto
-
Cross-Domain Image-Based 3D Shape Retrieval by View Sequence Learning
-
SeeThrough: Finding Objects in Heavily Occluded Indoor Scene Images
-
Robust Fitting of Subdivision Surfaces for Smooth Shape Analysis
Spotlight Session 2
Thu. 6/9 11:00 - 12:20. Session Chair: Philippos Mordohai
-
Calibrating Focused Light-Field Cameras Using Plenoptic Disc Features
-
Unsupervised Depth Estimation from Light Field Using a Convolutional Neural Network
-
Detail Preserving Depth Estimation from a Single Image using Attention Guided Networks
-
Structured Adversarial Training for Unsupervised Monocular Depth Estimation
-
Learning monocular depth estimation with unsupervised trinocular assumptions
-
Human-Centric Scene Understanding from Single View 360 Video
-
Constant Velocity 3D Convolution
-
FEATS: Synthetic Feature Tracks for Structure from Motion Evaluation
-
An Image-based Approach for 3D Reconstruction of Urban Scenes Using Architectural Symmetries
-
Spatiotemporal Modeling for Efficient Registration of Dynamic 3D Faces
-
Survey of Higher Order Rigid Body Motion Interpolation Methods for Keyframe Animation and Continuous-Time Trajectory Estimation
-
Semantic Classification of 3D Point Clouds with Multiscale Spherical Neighborhoods
-
Efficient Convolutions for Real-Time Semantic Segmentation of 3D Point Clouds
-
Logistic Regression of Point Matches for Accurate Transformation Estimation
-
Analyzing body fat from depth images
-
Adversarial Semantic Scene Completion from a Single Depth Image
-
3DTNet: Learning Local Features using 2D and 3D Cues
-
Structured Low-Rank Matrix Factorization for Point-Cloud Denoising
-
Generative Adversarial Frontal View to Bird View Synthesis
-
Multi-View Inpainting for RGB-D Sequence
Poster Session 2
Thu. 6/9 14:00 - 16:00.
All the papers presented on 6/9
Oral Session 2.P
Thu. 6/9 16:00 - 17:15. Session Chair: Adrian Hilton
-
Rethinking Pose in 3D: Multi-stage Refinement and Recovery for Markerless Motion Capture
-
Neural Body Fitting: Unifying Deep Learning and Model Based Human Pose and Shape Estimation
-
3DBodyTex: Textured 3D Body Dataset
-
3D Head Pose Estimation Using Tensor Decomposition and Non-linear Manifold Modeling
-
Hybrid Skeleton Driven Surface Registration for Temporally Consistent Volumetric Video
Keynote Session 3
Fri. 7/9 9:00 - 9:45. Session Chair: Umberto Castellani
-
Geometric deep learning
Oral Session 3.A
Fri. 7/9 10:15 - 11:00. Session Chair: Sudipta Sinha
-
Fine-Level Semantic Labeling of Large-Scale 3D Model by Active Learning
-
UnrealStereo: Controlling Hazardous Factors to Analyze Stereo Vision
-
Learning a Hierarchical Latent Variable Model of 3D Shapes
Spotlight Session 3
Fri. 7/9 11:00 - 12:20. Session Chair: Jean-Sebastien Franco
-
The Double Sphere Camera Model
-
DSR: Direct Self-rectification for Uncalibrated Dual-Lens Cameras
-
Monocular Depth Estimation via Deep Structured Models with Ordinal Constraints
-
ESPReSSo: Efficient Slanted PatchMatch for Real-time Spacetime Stereo
-
Unsupervised Adversarial Depth Estimation using Cycled Generative Networks
-
TwinFusion: High Framerate Non-Rigid Fusion through Fast Correspondence Tracking
-
Learning to Reconstruct Texture-less Deformable Surfaces
-
Multi-Planar Monocular Reconstruction of Manhattan Indoor Scenes
-
Novel Single View Constraints for Manhattan 3D Line Reconstruction
-
Learning to Generate Facial Depth Maps
-
Accurate Sparse Feature Regression Forest Learning for Real-Time Camera Relocalization
-
ACE: An Efficient Asynchronous Corner Tracker for Event Cameras
-
A Data-driven Prior on Facet Orientation for Semantic Mesh Labeling
-
Parsing Geometry using Structure-Aware Shape Templates
-
A 3D Coarse-to-Fine Framework for Volumetric Medical Image Segmentation
-
Deep Learning Parametrization for B-Spline Curve Approximation
-
Reconstruction of interior walls from point cloud data with min-hashed J-linkage
-
Self Functional Maps
-
Residual MeshNet: Learning to Deform Meshes for Single-View 3D Reconstruction
Poster Session 3
Fri. 7/9 14:00 - 16:00.
All the papers presented on 7/9
Oral Session 3.P
Fri. 7/9 16:00 - 17:15. Session Chair: Olaf Hellwich
-
PCN: Point Completion Network
-
Weakly supervised learning of indoor geometry by dual warping
-
EOE: Expected Overlap Estimation over Unstructured Point Cloud Data
-
NRGA: Gravitational Approach for Non-Rigid Point Set Registration
-
DUGMA: Dynamic Uncertainty-Based Gaussian Mixture Alignment
Demo session
Wed.5/9 - Fri. 7/9. Session Chair: Riccardo Spezialetti
-
Towards real-time monocular and unsupervised depth estimation on CPU
Real-Time Monocular 4D Face Reconstruction using the LSFM models
-
Online Sparse Scene Coordinates Learning for Real-Time Camera Relocalization
Exhibition
Wed.5/9 - Fri. 7/9.
-
Agisoft
-
Aquifi
-
Google
-
Microsoft
-
Microtec
-
3Dflow
-
DigitalViews
Tutorials full-day
Sat. 8/9 h 9:00 - 17:30. Tutorials will be held at the University of Verona, Strada le Grazie 15, Verona.
- Room G:
Methods for photographic radiometry, modeling of light transport and material appearance
- Room "Verde":
Tutorials half-day
Sat. 8/9 9:00 - 12:30. Tutorials will be held at the University of Verona, Strada le Grazie 15, Verona.
Sat. 8/9 14:00 - 17:30. Tutorials will be held at the University of Verona, Strada le Grazie 15, Verona.
Presentation Instructions
Oral presentations
- Oral presenters (except spotlights) will use their own laptop.
- The projector will be setup for the 4:3 format using a 1024x768 resolution.
- The projector will have only a VGA connector.
- Oral talks are given 12 minutes. Since the sessions are packed, this will be strictly enforced. You must leave the podium once your time is up.
- Additional 3 minutes are allocated for questions by the session chairs, switching between speakers and introducing the next speaker.
- Oral presenters are also required to present a poster during the poster session in the same day as the oral presentation.
Spotlight presentations
- All papers accepted as poster are also presented in the “spotlight” session.
- Spotlight presenters are required to send a video recording of their slides in MP4 format using a 4:3 format at a resolution of 1024x768.
- Higher resolution are allowed, but keep in mind that the video will be projected at 1024x768 resolution.
- The video will be played from the control room, so there is no need to connect your laptop.
- The audio channel from your submitted video will NOT be used: you are still required to present the slides orally as your video plays.
- The recording must not exceed 3 minutes and 50 seconds. Longer videos will be truncated. The next speaker’s video will start at exactly 4 minutes. Timing will be strictly enforced.
- The session chairs will not be introducing each spotlight. Thus, start the video with a title slide and introduce yourself.
- Create a single MP4 file named PAPERNUMBER.mp4 where PAPERNUMBER is your 3DV paper number.
- Put the file in a place where it is sharable (e.g. Dropbox, OneDrive, FTP server, etc…) and send the link by e-mail to 3dv18demos@3dv.org
- The deadline is August 10 at 12:00 CET.
Poster sessions
- Poster stands have a dimension of 135 (width) x 100 (height) cm. You can use any poster size within these bounds, but we recommend using a horizontal A0 layout.
- Before the poster session begins, please identify the number of your poster in the program and attach the poster to the corresponding stand. Check with the volunteers or the registration desk if you cannot find your poster stand.
- After the poster session ends, please remove the poster to leave the space for the next presenter.