//Autonomous Driving
Autonomous Driving2019-08-20T18:12:51+00:00

Autonomous Driving

Synthetic Training & Validation Data

BIT provides training data sets as ground truth for machine learning and validation with highly scalable scene generation from real world scenarios. Different sensor outputs as well as corner case variations are produced with a high degree of automation. AI based & traditional methods produce scenes for this virtual simulation data.

BIT provides a virtual sensor simulation framework fully linked to ISO 26262 ASIL levels. Critical corner cases as well as automatic FMEDA output are processed.

bounding boxes

Multiple sensors and different output formats are available and can be adapted quickly on demand. We generate content for camera sensors, radar (in preparation) and laser/LIDAR. Optical flow and sceneflow analysis is available on request.

urban szenario The BIT digital content framework allows for automatic generation and randomized variation of scene aspects and behavior models.

We provide one of the largest industry sets for training data. Different production tools such as Unreal Engine™ as well as own render engines are used.

Real World Labeling

BIT takes pride in having a proven history of image labeling and scene annotation for different car OEMs. We are a key supplier for labeling 2D, 3D bounding boxes as well as annotation based on combined inputs such as LIDAR, radar and camera. Data protection compliant according to the EU General Data Protection Regulation.

Semantic Segmentation

Self-learning algorithm for automatic processing of camera data in unleashed quality. Delivery of segmented pictures and meta data in open JSON/XML field formats.

Synthetic dataset requirement specifications

Scene Flow

The Scene Flow is offered in different file formats and as colored image for reference checks. We support open file formats for easy adoption into existing tools. Sub pixel accuracy along with selectable resolutions are supported.

Depth

Along with the segmentation of the virtual scene we generate a depth map which is shown here a visual graphic.

Optical Flow

Independent of the Scene Flow an Optical Flow output engine is integrated in our pipeline allowing cross checks of the Scene Flow with same open file formats.

Localization / odometry
For cross check and implementation of odometry algorithms dedicated visiual odometry markers are implementable.

Semantic segmentation
Along different groups and color conventions such as City Scapes. Level of detailed up to each pixel with unique identifiers.

Freespace detection
BIT SW products detect and describe free space with a consisten file format which looks in a visual form like this:

Traffic sign detection

Growing catalog of international traffic signs & lights along with variations on effects such as dirty, damage, oclusion, bending, painting and others. We these variations critical corner cases are generated to increase the covergae of the learning.

Depht -Variant binary file

Depht- Variant ASCII text file

Standard pinhole camera

 

Fast adoption of synthetic models to offer corner case variations based on real world inputs. Easy adoption of camera parameters, resolution, dynamic car behavior to debug machine learning.

 

Inner city heavy traffic / congestion

Inner city
– Altheimer Eck + Variations and extensions

Highway

Germany / Switzerland / USA / Urban highway with divider / Highway entrance ramp and exit / Toll station / Mountain-road / Seaside / Forest / Desert

Parking open

Parking garage

 

Suburban
– inner and outer Parkring + variations and extensions

Normal / Night / Rain / Night with rain / Fog / Snow / Dawn/Dusk

NCAP scenarios / NCAP Emergency Brake Test

Jogger
– Animation can be done with behavioral models (in research with partners)

Pedestrian 1

Pedestrian 2

Child
Partly occluded objects / early detection of suddenly appearing objects

Partly occluded pedestrian

Partly occluded cyclist

Partly visible car / Bicycle

Child running onto street / objects falling onto the street

Detecting people’s poses/gestures

Cyclists giving hand signals for turning left/right / Policeman giving hand signals / Pedestrian behavior prediction

Unknown objects which weren’t trained in standard use cases

Animals (non stationary) / Unknown objects / obstacles (non stationary) / Unknown objects / obstacles (stationary)

“Soft” unknown objects / obstacles (stationary) – overdrivable / Objects hard to detect / Accidents

Static objects / content

Unknown static objects / obstacles

“Soft” unknown objects / obstacles – overdrivable

Dynamic objects

Car

Transporter/Caravan

Truck / Tram / Special vehicles / Trailer

Bicycles

Motorbikes

Pedestrians

Animals / Unknown dynamic objects

Urban

Traffic Calmed Zone / Residential Area

Bidirectional / One way

Sub Urban

Bidirectional / One way / With divider / Urban Highway / Urban Highway with divider

 

Single solid / Double solid / Dashed / Double dashed / Curb / Barrier Jersey / Guard Rail / Fence / Cliff / Ditch

Need more information?

Contact Us

In order to optimize our website for you and to be able to continuously improve it, we use cookies. By continuing to use the website, you agree to the use of cookies. Further information on cookies can be found in our privacy policy. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close