In spatio-temporal reasoning, granularity is one of the factors to be considered when aiming at an effective and efficient representation of space and time. There is a large body of work which addresses the issue of granularity by representing space and time on a qualitative level. Other approaches use a predefined scale which implicitly determines granularity (e.g., seconds, minutes, hours, days, month, etc.). However, there are situations where the right level of granularity is unknown in the beginning, and is only determined in the problem solving process itself. This is the case in machine learning, where the learner has to find a representation for a problem and with that the right granularity for representing space and time. This paper introduces an algorithm which determines the most appropriate level of granularity during training. It uses several description logic learners as the learners, and the positive and negative examples presented to them as the determinators for refining coarse temporal representations to the most appropriate level of granularity.

This paper aims at providing a preliminary discussion on how to deal with spatio-temporal information in the context of behaviour recognition. It draws comparison with how humans reason in other areas, such as law, and discusses some of the pros and cons of formalisms for handling uncertainty, starting with probability theory, continuing with the Dempster-Shafer theory, and concluding with fuzzy logic.

The cognitive interpretation of perceptual data (e.g., from video, depth, motion sensors) requires the representational and inferential mediation of commonsense and qualitative abstractions of space, actions, events, change, and interaction. General methods and benchmarks for high-level cognitive interpretation, and their seamless integration and access within large-scale projects concerned with cognitive vision, robotics, hybrid-intelligent systems are necessary. We present the Rotunde initiative as a particular instance of a challenging smart meeting cinematography concept primarily concerning human activity interpretation. The Rotunde initiative aims to release general tools (e.g., for reasoning and control), methodological and performance benchmarks, and developmental aids (e.g., management and visualisation of complex spatio-temporal data) for the cognitive interpretation of interaction.

This workshop has a special focus on the topic of spatio-temporal aspects of human activity interpretation, especially welcoming research concerned with monitoring and inter- pretation of people interactions, real-time commonsense situational awareness involving aspects such as scene perception and understanding, perceptual data analytics, and prediction and explanation-driven high-level control of autonomous systems. In this context, basic topics deemed important include activity and process models; behaviour and intention interpretation; spatial learning; modeling and reasoning about space, events, actions, interaction; spatio-temporal dynamics; and commonsense reasoning about spatio-temporal change.

In order to fully utilize robots for our benefit and design better agents that can collaborate smoothly and naturally with humans we need to understand how humans think. My goal is to understand the mirroring process and use that knowledge to build a computational cognitive model to enable a robot/agent to infer intentions and therefore collaborate more naturally in a human environment.

In exploratory domains, agents' actions map onto logs of behavior that include switching between activities, extraneous actions, and mistakes. These aspects create a challenging plan recognition problem. This paper presents a new algorithm for inferring students' activities in exploratory domains that is evaluated empirically using a new type of flexible and open-ended educational software for science education. Such software has been shown to provide a rich educational environment for students, but challenge teachers to keep track of students' progress and to assess their performance. The algorithm decomposes students’ complete interaction histories to create hierarchies of interdependent tasks that describe their activities using the software. It matches students' actions to a predefined grammar in a way that reflects that students solve problems in a modular fashion but may still interleave between their activities. The algorithm was empirically evaluated on people’s interaction with two separate software systems for simulating a chemistry laboratory and for statistics education. It was separately compared to the state-of-the-art recognition algorithms for each of the software. The results show that the algorithm was able to correctly infer students' activities significantly more often than the state-of-the-art, and was able to generalize to both of the software systems with no intervention.

We present a robust framework for complex event recognition that is well-suited for integrating information that varies widely in detail and granularity. Consider the scenario of an agent in an instrumented space performing a complex task while describing what he is doing in a natural manner. The system takes in a variety of information, including objects and gestures recognized by RGB-D and descriptions of events extracted from recognized and parsed speech. The system outputs a complete reconstruction of the agent’s plan, explaining actions in terms of more complex activities and filling in unobserved but necessary events. We show how to use Markov Logic (a probabilistic extension to first order logic) to create a theory in which observations can be partial, noisy, and refer to future or temporally ambiguous events; complex events are composed from simpler events in a manner that exposes their structure for inference and learning; and uncertainty is handled in a sound probabilistic manner. We demonstrate the effectiveness of the approach for tracking cooking plans in the presence of noisy and incomplete observations.

Referring expressions such as “a long meeting” and “a restaurant near my brother’s” depend on information from the context to be accurately resolved. Interpreting these expressions requires pragmatic inferences that go beyond what the speaker said to what she meant; and to do this one must consider the speaker’s decisions with respect to her initial belief state and the alternative linguistic options she may have had. Modeling reference generation as a planning problem, where actions corre- spond to words that change a belief state, suggests that interpretation can also be viewed as recognizing belief- state plans that contain implicit actions. In this paper, we describe how planners can be adapted and used to interpret uncertain referring expressions.

While recent Machine Learning (ML) based techniques for activity recognition show great promise, there remain a number of questions with respect to the relative merits of these techniques. To provide a better understanding of the relative strengths of contemporary Activity Recognition methods, in this paper we present a comparative analysis of Hidden Markov Model, Bayesian, and Support Vector Machine based human activity recognition models. The study builds on both pre-existing and newly annotated data which includes interleaved activities. Results demonstrate that while Support Vector Machine based techniques perform well for all data sets considered, simple representations of sensor histories regularly outperform more complex count based models.

In spite of the significant work that has been done todiscover and recognize activities in the smart home re-search, less attention has been paid to predict the futureactivities that the resident is likely to perform. An ac-tivity prediction module can play a major role in designof a smart home. For instance, by taking advantage ofan activity prediction module, a smart home can learncontext-aware rules to prompt individuals to initiate im-portant activities. In this paper, we propose an activityprediction approach using Bayesian networks. We pro-pose a novel two-step inference process to predict thenext activity features and then to predict the next activ-ity label. We also propose an approach to predict thestart time of the next activity which is based on model-ing the relative start time of the predicted activity usinga continuous normal distribution and outlier detection.We evaluate our proposed models using real data col-lected from two smart home apartments.