Time Table: Applied Vision @ AIRoV 2024

Wednesday March 27, 2024

Invited Talk

Cobotic Induction System for Garment Picking

Fabio Pini, PhD Researcher On Integrated Design Methodologies For The Development Of Industrial/Collaborative Robotic Solutions; University Of Modena And Reggio Emilia The growing trend of garments waste highlights the need of effective and affordable processes to enable a more sustainable lifecycle. Collaborative robotic systems that support operators represent a feasible solution, but there are still some limitations related to the manipulation of deformable objects, such as garments. By integrating versatile gripper system and machine learning models trained to detect each garment, this work suggests a solutions to automate garments picking. The first experiments highlight interesting results about effective grasping and promising further improvements about the productivity.

Schedule

First Session 9:00-10:30

Chair: Applied Vision Team

• 09:00-09:05 Introduction
• 09:05-09:30 Invited Talk
• 09:30-09:45 Camera-based High-Speed Rolling Mark Detection (ID: 10) - Manuel Krammer (Plasser & Theurer, Export von Bahnbaumaschinen, GmbH)*
• 09:45-10:00 Initial Results on Hyperspectral Characterization of Steel Scrap (ID: 48) - Malte P Jaschik (Joanneum Research)*; Harald Ganster (Joanneum Research); Maria Kainz (Joanneum Research); Maria Jernej (Joanneum Research)
• 10:00-10:15 Robustness of One-Class Visual Anomaly Detection Methods (ID: 45) - Florian Kromp (Software Competence Center Hagenberg GmbH)*; David Brunner (SCCH); Johannes Zellinger (Software Competence Center Hagenberg GmbH); Alexander Valentinitsch (Software Competence Center Hagenberg GmbH); Ozan Cakiroglu (Software Competence Center Hagenberg GmbH); Clara Schachner (Software Competence Center Hagenberg GmbH); Lukas Fischer (SCCH)
• 10:15-10:30 Simulating Human Perception: Crafting Datasets for Advanced Print Defect Detection on Packaging Foil (ID: 47) - Lukas Brunner (Profactor)*; Thomas Pönitz (Profactor)

Second Session 11:00-12:30

Chair: Applied Vision Team

• 11:00-11:15 Transparency Techniques for Neural Networks trained on Writer Identification and Writer Verification (ID: 25) - Viktoria Pundy (TU Wien)*; Marco Peer (TU Wien); Florian Kleber (TU Wien)
• 11:15-11:30 Parameter Identification for Pattern-Generating Reaction-Diffusion Systems – Towards Generative Texture Descriptors (ID: 17) - Werner Ainhauser (UMIT Tirol); Martin Welk (UMIT TIROL)*
• 11:30-11:45 Supporting Image-Based Wildlife Classification (ID: 1) - Sead Mustafic (Joanneum Research); Dominik Dachs (Meles Wildbiologie); Rainer Prüller (Pentamap); Florian Schöggl (Pentamap ); Roland Perko (Joanneum Research)*
• 11:45-12:00 Solving Microorganism Enumeration through Weakly-Supervised Counting with Vision Transformers - A comparative study (ID: 27) - Javier Ureña Santiago (University of Innsbruck)*; Antonio J Rodriguez-Sanchez (University of Innsbruck)
• 12:00-12:15 Enhancing Historical Image Retrieval with Compositional Cues (ID: 43) - Tingyu Lin (TU Wien)*; Robert Sablatnig (TU Wien)
• 12:15-12:30 Advanced Despeckling Based on Real SAR Intensity Images (ID: 6) - Katharina Hofer-Schmitz (Joanneum Research); Sead Mustafic (Joanneum Research); Karlheinz Gutjahr (Joanneum Research ); Roland Perko (Joanneum Research)*