Whaaat the hack?

Why do we need a Workshop on Computer Vision Challenges in Industry?

Over the last years, the world-wide computer vision and machine learning community has grown significantly. Within this growing field, the academic research institutions of German-speaking countries play a substantial and still increasing role, which can be easily seen in the number of international collaborations and the number of papers in highest-ranked venues. Despite this strong competence, the participation of industrial computer vision researchers from German-speaking countries on *the* local conference in the field, the German Conference on Pattern Recognition (GCPR), is scant.

As one consequence, the exchange between academia and industry in German-speaking countries is substantially weaker than in other countries, e.g., compared to the US. This lack of exchange happens both on the level of content (i.e., comparably few joint research projects) and on the level of personal exchange (e.g., rarely happening internships during PhDs). However, both sides would greatly benefit from an improved exchange, e.g., by having joint research projects with real-world impact, by enabling reliable career paths for students, or by drawing inspiration from regular discussions on the boundary of fundamental and applied research. This workshop tries to bridge the gap between results of basic research on computer vision and their real-world applicability with a joint exchange forum. Thereby, the workshop aims at reaching the following goals:

  • Bring together researchers from industry and academia working on pattern recognition, machine learning, and computer vision and its applications
  • Highlighting recent challenges in industrial research to demonstrate the potential impact of upcoming fundamental research
  • Exchange best-practices, brave new ideas, and shortcomings when transferring academic research results into industry projects
  • Fostering collaborations among PIs from academia and industry


01:30pm - 01:35pm
Welcome and Introduction
01:35pm - 02:00pm
Christian Wojek (ZEISS) - Computer Vision for Microscropy Applications
02:05am - 02:30pm
Nils Hasler (The Captury) - Computer Vision Challenges at StartUps
02:35pm - 03:00pm
Tobias Klinder (Philips) - Computer Vision and Machine Learning Challenges in Diagnostic Imaging
03:05pm - 03:30pm
Carsten Steger (MVTec) - Computer Vision for Industrial Applications
03:35pm - 04:00pm
Uwe Franke (Daimler) - Computer Vision for Automotive Applications
04:05pm - 04:30pm
Panel discussions and questions to speakers
Workshop Closing

Invited speakers - details

Christian Wojek (ZEISS)
Christian will speak about challenges in using computer vision for microscopy applications.
Nils Hasler (The Captury)
As anyone knows who has worked in both worlds, topics interesting for academics may differ from topics that are applicable in the real world. In my presentation I will touch on some topics where academia and industry disagree on what's interesting and useful. Nils will specifically focus on challenges in using computer vision in StartUps.
Tobias Klinder (Philips)
With the pressure on health systems to provide better outcomes at lower costs, it is increasingly becoming evident that Computer Vision and Machine Learning will play a pivotal role in medical applications. However, in order to offer value-adding solutions, understanding of the market is crucial. In this talk, I will show specific examples to illustrate the challenges in the diagnostic imaging industry and to discuss the different focus of industrial versus academic research.
Carsten Steger (MVTec Software GmbH)
Machine vision is a crucial technology in the automation of many industrial processes. Carsten will speak about challenges faced by users in the machine vision industry, with a focus on quality control and inspection.
Uwe Franke (Daimler)
Computer Vision became a key for driver assistance as well as for future autonomous vehicles. For nearly 30 years Uwe focused on research and application in this field. He will address inspirations that have been triggered by academia and enabled solutions in industry. In addition he will dive in challenges ahead.


The following topics will be covered in the workshop:

  • Presentation of recent industrial challenges in established computer vision fields, such as classification, segmentation, novelty detection, or 3D object detection
  • An overview on most pressing problems from an application perspective
  • Problems and solutions for dealing with small amounts of data in real-world applications
  • Discussing future trends in industry and where the community can contribute
No matter if you're a student, PhD student, PostDoc, group leader, PI, industrial researcher, or project team lead - we invite you to participate!