Data Coding Using Sparse Mixture of Experts Regression

16002/TUB: Data Coding Using Sparse Mixture of Experts Regression


The primary goal of the method described here is to allow the reconstruction of images in high quality by using a universal image coder with easy bit-level access to MPEG-7-like low- and mid-level image features in the decoder. Natural images are mostly piecewise smooth. Therefore, the idea is to search for unsteady regions in the image, and to approximate the stationary regions separately, but smoothly combining both of them with great care.

Technical Description

The technique presented here offers a universal, bit-efficient video compression approach with high MPEG-7-like features intrinsically embedded in the bitstream. A sparse Mixture-of-Experts (MoE) regression approach for encoding videos in the pixel domain is used, deferring drastically from the established DPCM / Transform coding philosophy. Here the MoE takes the form of a Gaussian Mixture Regression (GMR) for multivariate nonlinear regression. The underlying stochastic process of the pixel amplitudes are modelled as a 3-dimensional and multi modal mixture of Gaussian modes with K modes. Therefore, each component in the MoE steers in the direction of the highest correlation.

Experiments show that – compared to JPEG – for a large class of images a considerable compression gain can be achieved at low bitrates, while providing attractive low-level descriptors for the image. In this way the SMoE shows a strong resemblance to MoE neural networks, while providing a performance that can compete with H.264.

a) Original                                      b) JPEG                                       c)SMoE

Anwendungsmöglichkeiten / Possible applications:

The method provides an almost universal bit-efficient video compressing approach that can be used in various coding applications.

Interested? Contact Us! Interested? Contact Us!
Show all techoffers Vorteile / Benefits:
  • Universal Image/Video Coding
  • Compression gain at low bitrates
  • MPEG7-like features embedded
  • Performance competitive to H.264
Entwicklungsstand / Development status:

lab level

Reifegrad / Maturity level:


Schlagworte / Tags:
bit-efficient, Image compression, steered Mixture-of-Experts, video coding
Schutzrechte / Property rights:

PCT - pending

Patentinhaber / Patent holder:

Technische Universität Berlin

Möglichkeiten der Zusammenarbeit / Possible cooperation:
  • R&D Cooperation
  • Patent Purchase
  • Licensing

Ansprechpartnerin / Contact Person:
Ina Krüger

Tel.: 030 314-75916
Zentrum für geistiges Eigentum