Efficient Recording and Playback of Haptic Interaction
Final Report Abstract
Over the past decade, haptics (the sense of touch) research has been gaining in relevance in both the multimedia and robotics communities. The most interesting aspect of haptic communication is that it is inherently bidirectional, i.e., humans not only feel haptic feedback - similar to audio/video - but also, physically act upon an environment. It is widely accepted that the communication of haptic data over networks has a far-reaching impact on important applications like on-orbit servicing for satellites, space exploration, tele-surgery, deep-sea exploration, safety-critical situations, tele-manufacturing, tele-manipulation and tele-assembly. Haptics also plays a key role in virtual simulations of real (e.g., collaborative assembly/design) or fantasy environments (e.g., games). The general topic that has been researched under this project is the development of efficient and perceptually transparent compression schemes for the transmission/storage of haptic data with a special focus on tactile information. Since the final sink of haptic media is a human operator, the limitations of human haptic perception have been leveraged to achieve compression. For haptic kinesthetic signals (e.g., force), the Weber’s law of psychophysics has been used for this purpose. On the other hand, for haptic texture signals, frequencydomain masking phenomena have been successfully exploited for compression. In this project we have developed two versions of a tactile compression engine. The first one operates at a constant bitrate (4 kbps) with a compression ratio of 8:1. This codec has been shown to be perceptually transparent through extensive subjective tests. Furthermore, we have extended this codec for bitrate-scalability with a bitrate range from 2.3 to 4 kbps. Preliminary subjective evaluations show that even at 2.3 kbps, the perceptual transparency is maintained. Another equally important aspect of the project has been the development of haptic action and perception models. This aspect has the dual purpose of serving as a basis for data compression algorithms, and for algorithmically predicting haptic signal quality in a model-based manner, thus circumventing the constant need for difficult and time-consuming subjective tests in haptic system evaluations. We have proposed the first objective quality metric which incorporates human action and perception models. The quality prediction given by these models qualitatively matches with the results of corresponding subjective tests.
Publications
- Opening the haptic loop: Network degradation limits for haptic task performance, Haptic Audio Visual Environments and Games (HAVE), 2011 IEEE International Workshop on, pp.56-61, Qinhuangdao, Hebei, China, 14-17 Oct. 2011
R. Chaudhari, C. Schuwerk, V. Nitsch, E. Steinbach, B. Färber
- Towards an objective quality evaluation framework for haptic data reduction, World Haptics Conference (WHC), 2011 IEEE, pp.539-544, Istanbul, Turkey, June 2011
R. Chaudhari, E. Steinbach, S. Hirche
- Haptic Communications, Proceedings of the IEEE, vol. 100, no. 4, pp. 937- 956, April 2012
E. Steinbach, S. Hirche, M. Ernst, F. Brandi, R. Chaudhari, J. Kammerl, I. Vittorias
- Low Bitrate Source-filter Model Based Compression of Vibrotactile Texture Signals in Haptic Teleoperation, Proc. of ACM Multimedia 2012, pp. 409-418, Nara, Japan, November 2012
R. Chaudhari, B. Cizmeci, K. Kuchenbecker, S. Choi, E. Steinbach
- Perceptually Robust Traffic Control in Distributed Haptic Virtual Environments, Proc. Eurohaptics Conference, Tampere, Finland, published in Haptics: Perception, Devices, Mobility, and Communication, Lecture Notes in Computer Science, Vol. 7282, pp 469-480, 2012
C. Schuwerk, R. Chaudhari, E. Steinbach
