Beamr Imaging Ltd. began in 2009 by the 3-time entrepreneur Sharon Carmel who has two previous startup exits to his credit which include a digital graphics and imaging technology company that scaled to a $7 billion valuation on the London stock exchange, and a cloud synchronization company acquired by Phoenix Technologies. Beamr serves the world's top Pay TV and OTT video service providers, the result of being a leading designer of content-adaptive video encoding and content-adaptive video optimization solutions that enable high quality, high performance, and bitrate efficiency for managed service operators, streaming video platforms, telcos, and online video streaming platforms.
The video technology Beamr developed for our codecs are covered by 11 granted patents that are a part of the 44 total patents the company holds that cover our breakthrough video optimization innovations. Though video quality and bitrate continues to be a core focus of development, Beamr recognized the prevailing trend to increase compute performance and density of software-based video encoding and transcoding infrastructure as a key to securing a competitive advantage for video distributors, streaming video platforms, and CDNs. At the same time, video system operators must enable encoding and delivery of advanced entertainment content, by embracing new technologies, capabilities, and codecs such as HEVC, HDR, and 4K.
Beamr's perceptual quality measure is based on a proprietary, low complexity, highly reliable, perceptually aligned quality metric that is covered by thirty-three International granted patents. The existence of this proprietary quality measure enables controlling a video encoder to output a stream with the maximum AMD Beamr compression possible while maintaining the original video resolution, format and visual quality. The way that this is achieved is by controlling the compression level of each frame in the video sequence in such a way that it is compressed to the maximum while producing a visual output that is the same as the quality that was originally specified.
The Beamr quality measure has been in development since 2009. After extensive refinement, and subjecting it to the strict requirements and testing of ITU BT.500 which is an International standard for certifying image quality, we were able to confirm that there is a sufficiently high correlation of our quality measure with human viewing results. The Beamr quality measure takes into account temporal artifacts. Since it is not sufficient to ensure only at the frame level that degradation has not been introduced - it is also necessary to preserve the quality of the video's temporal flow. When viewing a video encoded using the VBR rate-control and a video encoded with the CABR rate-control side-by-side, they will appear identical to the human eye under normal viewing conditions. This is a secret to how the process is able to guarantee image quality.
Massive advances have been made in software encoding, video codecs, and tools, yet the pace of video growth on the network is expanding faster than the availability of CPUs, causing silicon encoding solutions using GPUs, FPGAs, and ASICs to come back into focus. Beamr silicon accelerated CABR encoding opens new opportunities with very low TCO for mobile, edge, and cloud applications. The work Beamr is doing on silicon is closing the gap between the high performance of hardware encoders, and the complexity of advanced algorithms like CABR which traditionally were only able to be implemented in software on a CPU.