German Automaker Labels 12M Objects per Annum from Street Videos

Overview

The proliferation of sensors and cameras fosters next-generation ecosystems such as smart cities, smart transportation, smart infrastructure, etc. The value lies in accurately translating visual data into actionable insights in real-time. The customer wanted to automate the labeling process for street videos with live footage from various daytime conditions and locations to identify objects such as vehicles, street lights, and pedestrians.

Problem

Manually labeling objects from street videos was leading to:

Protracted cycles for object analysis
Increased cost due to higher manual labor
Higher scope for human errors

Solution

Apexon brought onboard its Image & Video Analytics platform, a proprietary deep learning platform, that speeds up analysis of visual data to enable:

Proactive prediction of objects seen in the video with deep learning algorithms such as EfficientDet, Resnet 50, Mask CNN, and the COCO weights
Curation of tags and corresponding images with XML parser - a data generator tool
Identification of object location across all frames and annotation generation with object tracking algorithms and computer vision
Conversion of annotations to XMLs with an XML generator to calibrate pinhole camera position and tool resolution

Impact

With our deep learning platform for visual data analysis, the customer was able to:

Drive 45% FTE savings with labeling automation
Improve the quality of annotation by 10-20%
Achieve more than 92% labeling accuracy
Reduce time to label new objects by 50% with reusable computer vision components

← BACK

Other Case Studies

US-based Physicians Group Improves ROI by 6X

Case Studies

Healthcare & Life Sciences

US-based Physicians Group Improves ROI by 6X

Learn how we helped a large physicians group based in the US use deep learning models to improve error identification by 3x and reduce financial leakage by 40%.

Learn More →

Automotive & Manufacturing

German Automaker Labels 12M Objects per Annum from Street Videos

Discover how we helped a German engineering firm automate video labeling for their autonomous vehicle project to identify street objects and their location with over 82% accuracy.

→

Financial Advisor Gets Retirement Benefits Calculator to Predict Over 30 Years Ahead

Case Studies

Financial Services & Insurance

Financial Advisor Gets Retirement Benefits Calculator to Predict Over 30 Years Ahead

Discover how we helped a financial advisor offering online investment advice evaluate the feasibility of their retirement plans 30 years into the future.

Learn More →