SEQUOIA: Sustainability-driven high quality video compression and delivery

The SEQUOIA project is a response to the emerging imperative needs for more financially and environmentally-sustainable high-quality video streaming. Project activities are aimed at addressing exploding market demands, and address the challenges faced by the media sector in tackling the **surge in online media consumption**, which is posing **unprecedented stress on Internet services**, while also greatly **contributing to the environmental footprint left behind by media services**.

Disruptive innovation for video streaming is urgently needed - new pre/post processing, encoding, and delivery tools that are device-aware and cross-codec compatible - to meet growing demands for online video reducing processing, energy and storage requirements. The SEQUOIA scientific innovation will result in a "**deep perceptual optimiser**" leveraging: (i) latest advances in psychovisual/perceptual models (ii) cutting-edge AI/deep-learning and (iii) state-of-the-art video coding, optimising quality, bitrate and environmental impact of video services.

Building on the partners' expertise on perceptual quality optimisation, video pre/post processing, video coding and deep learning, SEQUOIA will go beyond state-of-the-art to develop new AI-based video quality **analytics, streaming and encoding.** As a result, the SEQUOIA vision is to develop an automated **deep perceptual optimiser** to make an impact throughout all stages in the media distribution chain.

The project will demonstrate its results on operational and portable encoder designs, applicable both to Video on Demand (VOD) as well as live applications and encoding social media streams. The value of such scientific innovation will lead to benefits for the whole sector, demonstrating technology to enable sustainable distribution of Ultra High-Definition (UHD) content, while limiting the impact of video on internet traffic and reducing distribution costs. Extending beyond the commercial benefits, the project outcomes will be devised to support **environmentally conscious solutions by monitoring and proactively reducing energy consumption** at all stages within the media value chain.

Extending over a period of 24 months, the project brings together industrial (iSIZE and BBC) and academic (QMUL) partners. The partners bring unique expertise on **AI, video coding standardisation, adaptive video pre/post processing and streaming, perceptual optimisations and interoperable software architectures**, to collaboratively work towards these challenging objectives.