Workshops

TUCN

Workshop Name:
Deep Learning in Perception and Control

Organizers:
Sergiu Nedevschi, TUCN

Location: DoubleTree by Hilton Hotel Cluj – City Plaza, "Beijing" Room (5th floor)

Abstract:

This workshop represents an introduction in deep learning based solutions for autonomous systems. The workshop is based on the authors’ experience gain in several research projects carried out at TUCN dealing with perception and control tasks. The approached topics are deep learning based stereo reconstruction, object detection, panoptic segmentation, intent detection and slot filling and models interpretability.

Presentations:

1. "Real-time Simultaneous Monocular Depth Estimation and Object Detection”, Vlad Miclea, Sergiu Nedevschi
2. "Multi-task Network for Panoptic Segmentation in Self Driving Cars”, Andra Petrovai, Sergiu Nedevschi
3. "A Deep Learning solution for Intent detection and slot filling", Anda Stoica, Tibor Kadar, Rodica Potolea, Mihaela Dinsoreanu, Camelia Lemnaru
4. “Model Interpretability in Deep Learning”, Daniela Palcu, Marius Supuran, Mihaela Dinsoreanu, Camelia Lemnari, Rodica Potolea

BOSCH

Wokshop Name:
Bosch Students Workshop In Automated Driving
Organizers:
Catalin Golban, BOSCH

Location: DoubleTree by Hilton Hotel Cluj – City Plaza, "Beijing" Room (5th floor)

Abstract:

Video based driver assistance systems is a continuously growing field in automotive industry. Based on surround sensors such as radar, video and ultrasound, driver assistance systems sense and interpret the traffic scenes. They assist the driver in various driving situations and increase the driving comfort. In addition, driver assistance systems improve driving safety by supporting the driver in critical driving situations that require rapid and safe action. Gradually increasing performance will lead to automated driving solutions available in series cars in the upcoming years.

After an overview of Bosch activities in the video-based driver assistance systems and functions, the focus will be on a set of advanced video development topics .The workshop will present in detail and at tutorial level several advanced video algorithmic methods developed by Bosch R&D engineers in collaboration with students and researchers from Cluj universities in the last years, highlighting the importance of such methods in the context of the growing automated driving industry and showing the beauty and the challenges that appear when putting such methods on embedded hardware running in the car.

Presentations:

1. "Introduction and Overview of the Research Topics in Bosch Engineering Center Cluj", Catalin Golban
2. "Low Elevated Objects Detection for Autonomous Vehicles", Istvan Nagy
3. "Compression of Convolutional Neural Networks by Filter Pruning", Csanad Sandor
4. "Refraction Model Based Distortion Estimation for Video Sensors", Szabolcs Pavel
5. "Video Based Road Condition Monitoring", Paolo Moldovan