The areas of machine learning and communication technology are converging. Today’s communications systems generate a huge amount of traffic data, which can help to significantly enhance the design and management of networks and communication components when combined with advanced machine learning methods. Furthermore, in areas such as multimedia communication learning algorithms play an increasingly important role.
One part of our research is concerned with the development of efficient video analysis algorithms, which operate with the data encoded in compressed video bitstream such as motion vectors, block coding modes or transform coefficients of the motion-compensated prediction residuals. Compressed domain approaches generally have lower computational cost compared to pixel domain approaches since they avoid a full decoding of the video, thereby reducing the amount of processing and storage requirements significantly.
- W. Samek, S. Stanczak, and T. Wiegand, "The Convergence of Machine Learning and Communications", ITU Journal: ICT Discoveries - Special Issue 1 - The Impact of Artificial Intelligence (AI) on Communication Networks and Services, September 2017.
- V. Srinivasan, S. Lapuschkin, C. Hellge, K.-R. Müller, and W. Samek, "Interpretable human action recognition in compressed domain", Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2017), New Orleans, LA, USA, pp. 1692-1696, March 2017.
- V. Srinivasan, S. Gül, S. Bosse, J. T. Meyer, T. Schierl, C. Hellge, and W. Samek, "On the robustness of action recognition methods in compressed and pixel domain", Proceedings of the European Workshop on Visual Information Processing (EUVIP), Marseille, France, pp. 1-6, October 2016,
- S. Gül, J. T. Meyer, T. Schierl, C. Hellge, and W. Samek, "Hybrid Video Object Tracking in H.265/HEVC Video Streams", Proceedings of the International Workshop on Multimedia Signal Processing (MMSP), Montreal, Canada, pp. 1-5, September 2016.