Beamr Imaging Ltd. (NASDAQ: BMR) will demonstrate its machine learning-safe compression technology for autonomous vehicles at Vehicle Tech Week Europe, scheduled for June 23-25 in Stuttgart, Germany.

The video optimization company will present "From uncertainty to confidence: ML-aware video data for autonomous vehicles" on June 24 from 10:30-10:55 on the AVT Live Stage. The presentation will address how to integrate compression across autonomous vehicle technology stacks while maintaining machine learning model accuracy.

According to the company's benchmarks, Beamr's technology achieved up to 50% file size reduction with less than 2% difference in object detection accuracy for autonomous vehicle applications. For world foundation model pipelines, the technology delivered 41%-57% file size reduction with no measurable impact on captioning outputs.

"Working closely with AV engineering teams has given Beamr a clear view of where the real challenges sit - data volumes, model accuracy across the pipeline, integration into established workflows," said Dani Megrelishvili, Beamr Chief Product Officer.

The company's recent research showed that Beamr-compressed footage can strengthen depth models trained on the compressed data. The technology addresses accuracy and consistency metrics for petabyte-scale real-world fleet footage and synthetic data used in autonomous vehicle development.

Vehicle Tech Week Europe brings together original equipment manufacturers, Tier 1 suppliers, and autonomous vehicle development teams. Beamr will operate booth A256 during the event.

Based in Herzliya, Israel, Beamr develops content-adaptive video compression technology and holds 53 patents in the field. The company's clients include Netflix and Paramount, and its technology has received an Emmy Award for Technology and Engineering.