FAME @ IBC (cancelled)
AI-Powered Per-Scene Live Encoding
The world's Internet traffic is predominantly made up of video streaming, which can lead to 'bandwidth hogging' and loss in quality when large volumes of content are disseminated online. With an objective of achieving bitrate savings while maintaining optimal quality, our FAMIUM Deep Encode solution utilizes AI methods to enhance and automate Per-Title-Encoding and Per-Scene-Encoding processes.
FAMIUM SAND is a streaming analytics solution that collects and analyzes streaming metrics to provide in-depth video client session data and streaming performance information like average throughput, buffer level, representation switch events and initial playout delay (QoE metrics defined in ISO/IEC 23009-1) to help deeply understand streaming problems and client behavior to allow optimization of workflows.
FAMIUM DAI enables dynamic ad insertion (DAI) and flexible over-the-top streaming workflows including VoD-to-Live services across a multitude of devices and platforms via MPEG DASH and HLS. DAI can be delivered per user and in real time, in traditional live linear streams and on-demand content. FAMIUM DAI enables dynamic ad substitution (DAS) in HbbTV and supports SCTE35 based cue point signaling.
Since December 2019, Fraunhofer FOKUS maintains the source code of the dash.js project. Our work includes labeling, commenting, resolving and closing open issues and pull requests as well as implementing new features specified within the DASH-IF. The robustness, the wide support of features and the possibility of customizing the player makes dash.js an ideal client used in production by multiple companies and a large number of our own customers.