Semi-autonomous reference data generation for perception performance evaluation

Tatschke, T.; Farber, F.-J.; Fuchs, Ernst; Walchshausl, L.; Lindl, R.

doi:10.1109/icif.2007.4408000

Cited by 5 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate the performance of the approach a system for the semi-autonomous generation of reference data [17] has been used. This allows to determine error rates for false detections and missed detections.…”

Section: Discussionmentioning

confidence: 99%

Model-based detection and tracking of objects using a 3D-camera

Schindler

2010

2010 IEEE Intelligent Vehicles Symposium

View full text Add to dashboard Cite

In modern driver assistance systems the environment perception especially the detection of vulnerable road users plays an important role. The analysis of such traffic scenes demands reliable and robust information on objects and their position. In this context a 3D-camera, offers a promising concept in providing both lateral resolution and depth information to supply this task. This paper presents models and methods for the detection and the tracking of objects using the range data of a 3D-camera.For that purpose the depth information of a 3D-camera is used for the reconstruction of the traffic scenario. Special and adapted methods of the field of machine learning allow to analyze the re-projected structures in order to extract object measurements from the sensors raw data. Finally stochastic state estimation is applied to propagate object hypotheses taking into account the measurements. The proposed methodology facilitates the integration and support of standard environment perception techniques used in todays advanced driver assistance systems.

show abstract

Section: Discussionmentioning

confidence: 99%

Model-based detection and tracking of objects using a 3D-camera

Schindler

2010

2010 IEEE Intelligent Vehicles Symposium

View full text Add to dashboard Cite

show abstract

“…Point clouds are manually labeled and exploited further for referencing images. Finally, in [5], [6] software tools are created to facilitate and speed-up manual generation of reference data for 2D/3D object detection. The main advantage of manual labeling is the high accuracy of reference data created.…”

Section: Introductionmentioning

confidence: 99%

Towards a Reference Data Generation Framework for Performance Assessment of Perception Systems

Defraiteur

Rodriguez

Mittet

et al. 2019

2019 IEEE Intelligent Transportation Systems Conference (ITSC)

View full text Add to dashboard Cite

Sensors and their associated data fusion techniques, play a crucial role in Autonomous Vehicle (AV) decisionmaking applications. Accurately evaluate performance and reliability of the perception sources is an important task to be able to know the consistency of this data fusion. In this paper, a reference data generation framework for assessing perception sensors performances is proposed. Our approach relies on the complementary use of three data sources: a highly precise 3D map with semantic information, a High Density range finder sensor and a GNSS-RTK/INS localization unit. 3D map provides semantic knowledge of the environment and HD range finder precisely senses ego-vehicle surroundings. Finally, 3D map and HD scans are geometrically associated using positioning information in order to combine them and to infer reference data. Thorough experiments were conducted to evaluate and validate the proposed approach. As a proof of concept, performances of a LiDAR-based road plane detection method were evaluated, quantified and reported in terms of precision and recall.

show abstract