ACTRESS
Adaptive Video Control for Real-time Mobile Health Systems and Services
Over the past decade, there has been increased interest in the development of low-delay and high-quality medical video communications systems. The scope of the project is to provide methods for real-time end-to-end systems that dynamically adapt to time-varying channel state and end-user equipment while maximizing clinical video quality or guaranteeing a minimum, acceptable level of clinical video quality. The approach is to seek a jointly optimum solution based on multi-objective optimization that maximizes (clinical) video quality while minimizing bandwidth and encoding time. The latter is expected to expedite the adoption of m-Health medical video communications in standard clinical practice by preserving the quality thresholds for remote diagnosis and decision making. In this context, ACTRESS has set the following objectives:
- To address the inherent time-varying nature of wireless networks into exploiting how to reliably build multimedia applications and more particularly medical video communications
- To generate forward prediction models for (clinical) video quality that will be based on no reference VQA algorithms
- To exploit the full potential of the HEVC standard in creating dense encoding spaces, which will in turn lead to precise forward models generation and therefore accommodate smooth adaptation to the time-varying input constraints
- Funding source: Research Promotion Foundation
- Call: Restart 2016-2020
- Project number: -
- Total funding: 159,574
- Duration: 2018-2020