Jakob Doppler

Collecting complex activity datasets in highly rich networked sensor environments

We deployed 72 sensors of 10 modalities in 15 wireless and wired networked sensor systems in the environment, in objects, and on the body to create a sensor-rich environment for the machine recognition of human activities. We acquired data from 12 subjects performing morning activities, yielding over 25 hours of sensor data. We report the number of activity occurrences observed during post-processing, and estimate that over 13000 and 14000 object and environment interactions occurred. We describe the networked sensor setup and the methodology for data acquisition, synchronization and curation. We report on the challenges and outline lessons learned and best practice for similar large scale deployments of heterogeneous networked sensor systems. We evaluate data acquisition quality for on-body and object integrated wireless sensors; there is less than 2.5% packet loss after tuning. We outline our use of the dataset to develop new sensor network self-organization principles and machine learning techniques for activity recognition in opportunistic sensor configurations. Eventually this dataset will be made public.

Simulating the Potential Savings of Implicit Energy Management on a City Scale

According to statistics and future prospects in the next few years world-wide energy consumption will increase significantly. Therefore not only more energy efficient technologies but also more extensive energy saving concepts have to be realized. We have developed an rdquoimplicit interactionrdquo based power saving concept, which automatically schedules and controls energy consumers depending on the recognized activities of users. Moreover, dynamically schedulable power consumption loads are shifted in time, so as to make use of the cheapest energy prices without compromising user comfort. In order to calculate the potential savings of this concept we have implemented a simulation framework executing energy consumption models on a city scale, which allows for more complex scenarios than being restricted to some devices or buildings only. Both the framework architecture as well as large scale energy consumption simulations are presented. Simulation experiments give strong evidence for our implicit energy management concept as a promising source of energy savings.

Variability in Foot-Worn Sensor Placement for Activity Recognition

On-shoe acceleration and orientation sensors have revealed as a potentially powerful means for capturing aspects of human gait. The placement of sensors however has been done intuitively and mostly without quantitative evaluation of sensor positioning. Based on recorded signals of the five placement options sole, heel, toe-cap, instep and ankle we built SVM classifiers using orientation-based features and evaluate their performance on three activity classes level walking, going upstairs and going downstairs. Finally we present an approach to a placement-invariant classification model and discuss the benefit for a bipedal sensing setup.

RaPScoM: towards composition strategies in a rapid score music prototyping framework

Especially in the low-budget and amateur score music production workflow the triangular communication between editor, director and composer is constrained by limited resources, tight schedules and the lack of common domain knowledge. Often sound-a-like ideas and precomposed material end up as static temp tracks in the rough cut and thus limit the flexibility in the composition process. Our rapid score music prototyping framework RaPScoM aims at supporting the workflow with a toolchain for semi-automated and customizeable temp track generation by exposing a set of high-level parameters based on semantic movie annotation derived from movie clip analysis and a set of basic composition rules that could be derived from exemplary (score) music. In this paper we give an overview of the framework architecture and discuss technical details on our semantic movie annotation strategy and some core composition components.

Accessibility considerations in designing a layperson's application for score music description

The process of composing score music for a movie includes two different types of stakeholders: musical experts (a composer) and non-experts (a movie director, producer, editor etc.). These different preconditions often result in difficulties regarding the interaction and communication between the involved individuals. Here, a threefold approach was taken to address this problem: Extracting salient affective and semantic score music description parameters for musical non-experts; statistical clustering and modeling of the results and using them as ground truth for human and machine-based composition experiments; as well as identifying key factors of a human-computer interface capable of fostering non-expert musical creativity by employing a minimized set of intuitive design parameters.