Software HEVC video decoder: towards an energy saving for mobile applications

With the explosive growth of mobile video consumption over the Internet, delivering video at high quality while controlling the energy consumption of embedded decoding devices is becoming a primary concern. In this context, this paper demonstrates that tailored energy savings strategies, coupled with a video quality assessment protocol, have the potential to reduce the energy consumption of video decoding. We investigate on a real-time, optimized HEVC decoder, the relationship between the perceived mobile video quality and the energy consumption of the decoder. In addition to low level optimizations for ARM Neon platforms, two energy reduction methods named Approximate and Skipping have been investigated. Results show that energy savings of up to 20% can be achieved by using these methods with the same subjective perceived video quality. These subjective results confirm objective measurements, using PSNR and SSIM metrics, that depict a limited video quality degradation.