Monday, February 3rd
11:00 – 13:00 Arrival and registration
12:30 – 14:00 Lunch
14:00 – 14:10 Opening
14:10 – 15:00 Invited talk: Tatiana Tommasi
(title: Learning to Generalize with Self-Supervision)
15:00 – 15:30 Coffee Break
15:30 – 17:00 Session 1: Recognition, Segmentation and Biometrics
17:00 – 19:00 Free time: swimming
19:00 – 21:00 Dinner
21:00 – later Free time
Tuesday, February 4th
7:00 – 9:00 Breakfast
9:00 – 10:30 Session 2: 3D Computer Vision (1/2)
10:30 – 11:00 Coffee break
11:00 – 12:30 Session 3: 3D Computer Vision (1/2)
12:30 – 14:00 Lunch
14:00 – 15:30 Session 4: Video Analysis, Motion and Tracking
15:45 – 17:30 Social event: guided tour of the Rogaška glasswork
17:30 – 18:30 Free time: swimming
18:30 – 21:00 Dinner
Wednesday, February, 5th
7:00 – 9:00 Breakfast
9:00 – 10:30 Session 5: Vision Applications and Systems, Vision for Robotics
10:30 – 11:00 Coffee break
11:00 – 12:30 Session 6: Dataset and Evaluation
12:30 – 13:00 Awards, closing, CVWW2021
13:00 – 14:00 Lunch
Sessions
Session 1: Recognition, segmentation and biometrics (session chair: Franc Solina)
- On Learning Vehicle Detection in Satellite Video
Roman Pflugfelder*, Axel Weissenfeld, Julian Wagner - EPOS: Estimating 6D Pose of Objects with Symmetries
Tomas Hodan*, Dániel Baráth, Jiri Matas - Training end-to-end neural network for weakly labeled defect detection
Jakob Božič*, Danijel Skočaj, Domen Tabernik - Learning privacy-enhancing face representations through feature disentanglement
Marija Ivanovska, Blaž Bortolato*, Peter Rot, Peter Peer, Vitomir Štruc
Session 2: 3D Computer Vision (1/2) (session chair: Jiri Matas)
- MAGSAC++, a fast, reliable and accurate robust estimator
Dániel Baráth*, Jiri Matas, Jana Noskova, Maksym Ivashechkin - Novel Ways to Estimate Affine Transformations from Image Pairs
Levente Hajder*, Zoltán Pusztai, Nghia Le Minh - Homography Estimation from Images of a Vehicle-mounted Camera for Vertical and Horizontal Planes
István Gergő Gál*, Dániel Baráth, Levente Hajder - USACv20: robust essential, fundamental and homography matrix estimation
Maksym Ivashechkin*, Dániel Baráth, Jiri Matas
Session 3: 3D Computer Vision (2/2) (session chair: Danijel Skočaj)
- General 3D Room Layout from a Single View by Render-and-Compare
Sinisa Stekovic*, Friedrich Fraundorfer, Vincent Lepetit - Sphere-based LiDAR-Camera Calibration
Tekla Tóth*, Levente Hajder, Zoltán Pusztai - Refining corresponding image regions under multi-view constraints
Iván Eichhardt*, Dániel Baráth - Segmentation and Recovery of Superquadric Models using Convolutional Neural Networks
Jaka Šircelj*, Tim Oblak, Klemen Grm, Uroš Petković, Aleš Jaklic, Peter Peer, Vitomir Štruc, Franc Solina
Session 4: Video Analysis, Motion and Tracking (session chair: Matej Kristan)
- D3S – A Discriminative Single Shot Segmentation Tracker
Alan Lukežič*, Jiri Matas, Matej Kristan - Towards Data-driven Multi-target Tracking for Autonomous Driving
Christian Fruhwirth-Reisinger*, Georg Krispel, Horst Possegger, Horst Bischof - A new semi-supervised method improving optical flow on distant domains
Tomáš Novák*, Jan Sochman, Jiri Matas - movie2trailer:Unsupervised trailer generation using Anomaly detection
Orest Rehusevych*
Session 5: Vision Applications and Systems, Vision for Robotics
(session chair: Vaclav Hlavac)
- Kornia — an Open Source Differentiable Computer Vision Library for PyTorch
Edgar Riba, Dmytro Mishkin*, Daniel Ponsa, Ethan Rublee, Gary R Bradski - A water-obstacle separation and refinement network for unmanned surface vehicles
Borja Bovcon*, Matej Kristan - Learning to Predict Lidar Intensities
Patrik Vacek*, Otakar Jasek, Karel Zimmermann, Tomas Svoboda - Fungi Recognition: A Practical Use Case
Milan Sulc*, Lukas Picek, Jiri Matas, Thomas Jeppesen, Jacob Heilmann-Clausen
Session 6: Dataset and Evaluation (session chair: Roman Pflugfelder)
- Image Matching across Wide Baselines: From Paper to Practice
Yuhe Jin, Dmytro Mishkin*, Anastasiia Mishchuk, Pascal Fua, Jiri Matas, Kwang Moo Yi, Eduard Trulls - CNN-CASS: CNN for Classification of Coronary Artery Stenosis Score in MPR Images
Mariia Dobko*, Bohdan Petryshak, Oles Dobosevych - HOnnotate: A method for 3D Annotation of Hand and Objects Poses
Shreyas Hampali*, Mahdi Rad, Markus Oberweger, Vincent Lepetit - Practical high-speed motion sensing: event cameras vs. global shutter
Ondrej Holesovsky*, Vaclav Hlavac, Radoslav Skoviera, Roman Vítek