Motivation and Objectives
There is an increasing demand for robots and intelligent systems that co-exist with and support humans in their everyday life. A key to realize such robots and systems is reliable perception of humans. They can behave intelligently and effectively by reliably detecting humans and understanding their behaviors, identities, and physical/mental states.
Visual and sensor-based perception of humans has been an important research area in robotics, and its application area is expanding, accelerated by recent progress of machine learning algorithms and deep neural networks. This full-day workshop brings together researchers and industry engineers working on novel approaches and applications / practical applications of perception of humans.
The goal of the workshop is to recognize the state-of-the-art of human perception research and to discuss its future directions and potential application scenarios. The workshop consists of several invited talks by leading researchers and oral and poster presentations of contributed papers.
Research and application papers are solicited in, but not limited to, the following area topics:
- human detection and tracking
- person re-identification
- human pose estimation
- face detection and recognition
- gesture recognition
- human anomaly detection for care and service robots
- human behavior analysis and surveillance
- human-robot interaction and collaboration
- perception of humans for mobile robots and self-driving cars
- deep learning-based approaches to perception of humans
- systems and software tools for perception of humans
- human detection and recognition in/for industrial scenarios
Contents of the Workshop
| Type | Speaker/Title |
| Invited talk | Dr. Kai Arras (BOSCH) SPENCER and Beyond: Human detection, tracking and prediction in very dynamic environments |
| Invited talk | Dr. Francesco Nori (Google DeepMind) The geometric foundation of the dynamics of physical interaction and AnDy's roadmap towards proficient human-robot collaboration |
| Invited talk | Prof. Emanuele Menegatti (University of Padova) OpenPTrack: an open source infrastructure based on ROS and PCL for people tracking and human-machine interaction |
| Contributed paper | Kohari, Miura, Oishi CNN-based human body orientation estimation for robotic attendant |
| Contributed paper | Camara, Merat, Fox Filtration analysis of pedestrian-vehicle interactions for autonomous vehicle control |
| Contributed paper | Guidolin, Carraro, Ghidoni, Menegatti A Limb-based approach for body pose recognition using a predefined set of poses |
| Contributed paper | Castaman, Burke, Menegatti A novel user interface for OpenPTrack based on web technologies |
Organizers
| Jun Miura (Chair) |
| Department of Computer Science and Engineering, Toyohashi University of Technology |
| Emanuele Menegatti |
| Department of Information Engineering, University of Padua |
| Nicola Bellotto |
| Department of Computer Science, University of Lincoln |
| Stefano Ghidoni |
| Department of Information Engineering, University of Padua |
Additional Program Committee Members
| Marco Carraro |
| University of Padua |
| Kazunori Umeda |
| Chuo University |
| Serhan Cosar |
| University of Lincoln |
| Zhi Yan |
| University of Technology of Belfort-Montbeilard |