User-adaptive systems are a recent trend in technological development. Designed to learn the characteristics of the user interacting with them, they change their own features to provide users with a targeted, personalized experience. Such user-adaptive features can affect the content, the interface, or the interaction capabilities of the systems. However, irrespective of the specific source of adaptation, users are likely to learn and change their own behavior in order to correctly interact with such systems, thereby leading to complex dynamics of mutual adaptation between human and machine. While it is clear that these dynamics can affect the usability, and therefore the design of user-adaptive systems, scientific understanding of how humans and machines adapt jointly is still limited. The main goal of this project is to investigate human adaptive behavior in such mutual-learning situations. A better understanding of adaptive human-machine interactions, and of human sensorimotor learning processes in particular, will provide guidelines, evaluation criteria, and recommendations that will be beneficial for all project within SFB/Transregio 161 that focus on the design of user-adaptive systems and algorithms (e.g. A03, A07, projects in Group B). To achieve this goal, we will carry out behavioral experiments using human participants and base our empirical choices on the framework of optimal decision theory as derived from the Bayesian approach. This approach can be used as a tool to construct ideal observer models against which human performance can be compared. In particular, project C05 will address the following questions:

 

1. What are the determinants of mutual adaptation between an adaptable user and a user- adaptive system? Typically, problems arise when attempting to evaluate the quality of mutually adaptive human-machine technology, for instance through user’s subjective reports. Such reports often do not explain the outcome of the human-machine interaction behavior and are therefore unreliable, either as predictions of future usage or as indications of effective implementations. Such self-assessments are valuable, but only insofar as they are supported and extended through quantitative modeling. To overcome this limitation in the study of interaction between humans and adaptive technology, we will develop and test empirically a computational model of mutual human-machine adaptation. By combining an experimental with a compu- tational approach, we will provide a framework for bridging quantitative and qualitative methods, which is necessary for understanding the mechanisms underlying mutual human-machine adaptation.

2. How does mutual adaptation change based on the sensory modalities involved? We plan to further confirm the validity of the proposed model by exploring the dynamics of mutual adaptation between user and multimedia systems in increasingly complex scenarios. To do this, we will focus on the different sensory modalities targeted during interaction (primarily vision, audition, proprioception, and touch), and investigate the effects on the dynamics of mutual adaptation and its resultant perceptual and behavioral qua- lities. Understanding the unique characteristics of each modality is fundamental, since different modalities can potentially affect the adaptation dynamics in specific ways. In fact, many of the critical factors determi- ning adaptation (e.g. the nature of the sensory feedback, the prior knowledge involved, the predictability of possible changes that might occur) are modality specific. Moreover, different modalities also imply different levels of energy transfer between human user and technology. For example, there is no energy transfer with an audiovisual display. On the contrary, in situations where haptic feedback is provided, a user is physically coupled with the device and there is thus an energy transfer. This would allow the user to benefit from actively interacting with the system, but it could also affect the stability of the adaptation process.

3. Can mutual adaptation enhance immersion in an interactive virtual environment? By providing an application scenarios to be used as a testbed for user-adaptive systems, we will aim to understand the processes that can lead to the design of more efficient technology for enhancing immersion, thereby also supporting the research of projects focusing on immersive systems (e.g. C06, D04). It is well known that in the perception of a virtual world as typically rendered today there are often systematic biases present, such as in the perceived compression of depth or in the estimation of speed of ego-motion. Therefore we will ask whether, by using user-adaptive systems, it is possible to overcome such systematic distortions in human perception and action that limit immersion in virtual environments. To this end, we will implement adaptive features to virtual and augmented environments in order to investigate if, and under what circumstances, such features can improve the perceptual qualities of the rendered environments. It has been speculated that systematic distortions and perceptual conflicts also contribute to the well-known effects of simulator sickness. We will therefore investigate whether user-adaptive immersive systems are effective in reducing these effects.

4. Do the interaction capabilities and experiences learnt in an adaptive system generalize to real-life, non-adaptive scenarios? We expect that users will interact with adaptive immersive technology in a way that is more similar to behavior in a natural environment than non-adaptive equivalents. For this reason, such systems will provide a useful framework for understanding human behavior in the generalization towards real-life scenarios. As a final goal in our project, we will focus on understanding how interaction with an adaptive multimedia system affects human behavior in the context of in the wild experiments. More specifically, we are interested in two aspects: namely, how we can predict human behavior by looking at what happens in adaptive environments, and whether it is possible to understand mechanisms by which humans might generalize the skills learnt in a mutually adaptive environment to real-life scenarios.

 

Dises Projekt ist ein Teilproject des SFB TRR 161 gemeinsam mit den Unis Stuttgart, Konstanz und der LMU.