Perception and Scene Understanding

 

 

 

 

 


Group Leaders: Dr. rer. nat. Martin Lauer and Dr. Carlos Fernández López

One of the largest and most important areas of research in the context of autonomous driving is the domain of perception and scene understanding. In an abstract sense this domain contains all steps that are part of creating an environment model that can afterwards be used for planning the behavior of an autonomous car. This begins with the automated processing of all data perceived by the car's sensors like camera-images, RADAR-measurements or point clouds collected by a LiDAR. The processed data is coreferenced with each other based on a calibration between sensors and afterwads fused into one common environment model. Using techniques and algorithms from the domain of pattern recognition and machine learning it is even possible to accurately classify observed objects, traffic participants and elements from the environment to a defined set of classes and predict their behavior several seconds into the future.

 

Research topics:

 

Self-supervised learning

In self-supervised learning, supervisory signals are generated from unlabled data via pretext tasks. In a self-driving vehicle, the sensors, such as cameras and LiDAR, can be utilized to create training data without human intervention. By leveraging the vast amount of unlabeled data generated during normal driving, self-supervised learning algorithms can extract meaningful features and representations from the raw sensor inputs. Therefore, self-supervised learning can play a crucial role in improving the perception and decision-making capabilities of self-driving cars, contributing to their safety and reliability.

Contact: M.Sc. Royden Wagner

 

Environment Perception Under Low-Visibility Condition

Automated vehicles must be capable of precisely perceiving the environment despite low-visibility conditions. We analyze the effects of various visibility conditions (darkness, glare, rain, snow, fog) on the results of environment perception algorithms. Our aim is to develop novel preprocessing and domain adaptation approaches to improve the cross-domain robustness of object detection and tracking methods.

Contact: M.Sc. Kaiwen Wang

 

End-to-end perception and prediction

Object tracking and trajectory prediction are closely related components in the autonomous driving pipeline, which have been researched independently in past years. However, in the inference time, errors from the tracking module will reduce the prediction quality, called cascading error problem. A promising solution is end-to-end perception and prediction. Furthermore, the prediction module can conversely assist the tracking module to obtain better tracking performance. For the entire autonomous driving pipeline, the problem is even more severe. An end-to-end autonomous driving system will be investigated in the future.

Contact: M.Sc. Yinzhe Shen

 

Conditional behavior prediction

Recently, the focus of traffic agent position prediction has shifted from implicitly predicting individual agents independent from their surrounding to predicting agents conditioned on the behavior of surrounding agents. We propose to predict pair-wise distributions for all pairs in a scene. Then, marginal distributions for individual agents can be derived from those combined distributions if necessary, but primarily, those combined distributions can directly be analyzed for conditioned behavior modes.

Contact: Dr.-Ing. Florian Wirth

 

Probabilistic pedestrian prediction

In order to correctly react on people's behavior, a prediction of their probably future positions is needed. In this video you see how pedestrians could be predicted using an artificial neural network that was trained with trajectories of humans in traffic scenes. The green circles are corresponding to the minimum prediction horizon of 1 second, the red circles are corresponding the maximum prediction horizon of 4 seconds. With this prediction the ego vehicle can now brake for pedestrians even if they did not start crossing the road, yet.

Contact: Dr.-Ing. Florian Wirth

 

Simultaneous object tracking and shape estimation

Traditional object tracking methods approximate vehicle geometry using simple bounding boxes. We develop methods that simultaneously estimate object motion and reconstruct detailed object shape using laser scanners. The resulting shape information has benefits for the tracking process itself, but can also be vital for subsequent processing steps such as trajectory planning in evasive steering.

Contact: Dr. rer. nat. Martin Lauer or Dr. Carlos Fernández López

 

Vehicle Prediction

Since most interactions with other traffic participants occur between vehicles, the task of vehicle behavior prediction is a core task of scene understanding. To complete this task it is necessary to process all available information about the surroundings including but not limited to: Recognized obstacles, road geometry, other traffic participants, drivable area and traffic rules that are applicable to the current situation. The multitude of information necessary as well as the complexity of potential interactions between vehicles present a difficulty challenge in the prediction process.

Contact: M.Sc. Jannik Quehl

 

Realtime Environment Perception with Range Sensors

We aim to find algorithms to process observations from range sensors in real-time. Here, we make the assumption that the motion of traffic participants can be estimated w.r.t. to a common ground surface. We investigate methods to transfer range sensor observations from 3D space to the ground surface. The ground surface itself can be expressed as a dense 2D grid map with each cell referencing to 3D points. With this method, fast algorithms can be applied within the dense grid representation and easily transferred back to the 3D domain.

Contact: M.Sc. Marvin Klemp

 

Robust visual tracking in traffic scenarios

Object tracking is an essential part for behavior analysis and trajectory prediction in autonomous driving. Vision-based devices are able to provide rich information about observed environments.
Our aim is to develop novel approaches to track objects in light of powerful image features, which can deal with challenging scenarios such as occlusions and deteriorated visual conditions.

Contact: M.Sc. Yinzhe Shen