Back

Robust Object Detection and Classification Algorithms

Request Number N280679
Need details
Description

IncWell's partner seeks algorithms capable of providing rapid and reliable object detection and classification for usage in an automotive environment.

 

Background

In order for fully autonomous vehicles to transition from concept to practical reality, the vehicle must be capable to responding to the presence of other vehicles in the road. For this to be possible, there exists a need for computer systems able to detect and identify those other vehicles. Unfortunately, such processing can involve a high degree of processing time and power which makes it impractical for automotive usage.

 

IncWell's partner seeks high performance, rapid, robust, & cost efficient algorithms to provide object detection and classification utilizing image data provided by the on-board imaging sensors.

 

Key Success Criteria

 

The successful technology will provide the following performance with minimal latency and computational resources:

  • Range Specifications (first time vehicle appears in sensor range):
    • First detection range:
      • Day: <= 160 meters
      • Night <= 140 meters
    • Maximum range rate:
      • Day: 33 m/s
      • Night: 20 m/s
    • Range error:
      • 0m - 45m: < 6%
      • 45m - 90m: < 8%
      • Greater than 90m: < 9%
  • Identify the following vehicle types:
    • car
    • truck
    • emergency vehicle
    • motorcycle
  • True/False rate:
    • True positive: between 0 and 60m: > 99%
    • True negative: between 60m and 160m: > 97%
    • False positive: between 0 and 160m: < 1%
    • False negative: between 0 and 60m: < 1%

 

Possible Approaches

IncWell's partner is open to all approaches capable of providing the performance desired above.

 

Approaches not of interest

None.

Preferred Collaboration types
  • Joint Development
  • Technology Licensing
  • To Be Negotiated
Items to be submitted

When responding to this request, please be sure to provide a non-confidential description of the technology or solution you propose along with information regarding your alignment with the Key Success Criteria as well as your desired business relationship with IncWell's partner.

 

Point of Contact
Jonathan Jakischa
Area of Interest
  • Computer Software
  • Application Development
  • Business Software
  • Consumer Software
  • Data Analytics
  • Databases
  • Information Systems
  • Knowledge Management Systems
  • Programming & Design
  • Other
Organization
IncWell
Due Date
January 15, 2018
Attachments
This page does not have file attachments.
This page does not have file attachments.
Comments