Beamr's Benchmark Testing Validates ML-Safe Video Data Workflows for Autonomous Vehicles

Core Insights - Beamr Imaging Ltd. has demonstrated that its patented Content-Adaptive Bitrate (CABR) technology can achieve up to 50% storage reduction for autonomous vehicle (AV) video data while maintaining machine learning (ML) model accuracy [1][5][6] Group 1: Technology and Performance - The benchmark testing showed an average file size reduction of approximately 48% with less than 2% difference in mean Average Precision (mAP), a key metric for object detection reliability [5] - CABR technology was validated against industry-standard workflows using the PandaSet dataset, focusing on object detection tasks essential for AV perception systems [4][5] - The testing confirmed robust results across multiple quality metrics, including PSNR and LPIPS, indicating the effectiveness of CABR in maintaining video quality [5] Group 2: Industry Context and Applications - The testing addresses a significant challenge in AV development: managing large volumes of video data while ensuring fidelity for accurate ML performance [3] - AV systems generate hundreds of petabytes of footage, leading to high costs in budget and infrastructure, highlighting the need for efficient data management solutions [3] - Beamr's technology is applicable across high-growth markets, including media and entertainment, user-generated content, machine learning, and autonomous vehicles [6][7] Group 3: Upcoming Events - AV teams and developers are invited to meet Beamr's video data experts at CES 2026 in Las Vegas from January 6-9 to discuss the results and ML-safe AV pipelines [2]