Logo Techfak

devisto

Decoding Energy Visualization Tool

This online tool includes a large database of coded video bit streams including properties like distortion, rate, and decoding energy. You can choose different input sequences, codecs, and configurations and plot rate-distortion as well as energy-distortion curves. Play around with different streams to compare the performance of different codecs.

The decoding energy estimation tool (DENESTO) for single bit streams be found here.

Another hint: Most functionality can be found for the BlowingBubbles sequence (class D).

Select your bit stream:

The diagrams above show the rate-distortion (left) and the decoding-energy-distortion performance (right) of the chosen sequences. The distortion is given in YUV-PSNR, the rate in bit stream file size, and the decoding energy in Joule. The decoding energy was measured on the pandaboard using either FFmpeg (for all codecs) or HM-13.0 (only for HEVC coded sequences coded with 8 frames maximum). Note that comparing different codecs, a higher decoding energy does not necessarily mean that the decoding process is more complex in general (it highly depends on the implementation).

 

For some sequences, the decoding energy and decoding time (if not real-time constrained) was measured on several different devices using different software. Choose the device and software of interest and compare the required decoding energies for the sequence chosen above!

The following list gives the publications that this webpage is based on. The measurement setup is explained therein, and the values were used in the evaluation sections.

Decoding energies for various decoding systems:

C. Herglotz; D. Springer; M. Reichenbach; B. Stabernack; A. Kaup, "Modeling the Energy Consumption of the HEVC Decoding Process," in IEEE Transactions on Circuits and Systems for Video  Technology , vol.PP, no.99, pp.1-1

DERDO-optimized decoding energies:

C. Herglotz and A. Kaup, "Joint optimization of rate, distortion, and decoding energy for HEVC intraframe coding," Proc. IEEE International Conference on Image Processing (ICIP), Phoenix,  AZ, USA, 2016, pp. 544-548.

Decoding energies for various codecs:

C. Herglotz; Y. Wen; B. Dai; M. Kränzler; A. Kaup, "A Bit Stream Feature Based Model for Video Decoding Energy Estimation," Proc. Picture Coding Symposium (PCS), Nuremberg, Germany, 2016