Intelligent Behavior Utilizing Light-weight Diagnosis

Intelligence in its decisions is a trait that we have grown to expect from a cyber-physical system. In particular that it makes the right choices at runtime, i.e., those that allow it fulfill its tasks, even in case of faults or unexpected interactions with its environment. Analyzing how to continuously achieve the currently desired (and possibly continuously changing) goals and adapting its behavior to reach these goals is undoubtedly a serious challenge. This becomes even more challenging if the atomic actions a system can implement become unreliable due to faulty components or some exogenous event out of its control. In this paper, we propose a solution for the presented challenge. In particular, we show how to adopt a light-weight diagnosis concept to cope with such situations. The approach is based on rules coupled with means for rule selection that are based on previous information regarding the success or failure of rule executions. We furthermore present a Java-based framework of the light-weight diagnosis concept, and discuss the results obtained from an experimental evaluation considering several application scenarios. At the end, we present a qualitative comparison with other related approaches that should help the reader decide which approach works best for them.

The Importance of Traditional Sports into College Physical Education Based on Big Data Dynamic Programming Algorithm

Physical education teaching is conducive to the cultivation of students’ lifelong sports consciousness, which can improve students’ health and enhance their physique. In order to explore the importance of traditional sports based on big data dynamic programming algorithm into college physical education, the video action recognition and segmentation technology based on big data dynamic programming algorithm is designed. The complex actions in traditional sports teaching video are divided into a series of atomic actions with single semantics. The human action results are modeled according to the relationship between complex actions and atomic actions, and the actions are completed, and the changes of students’ sports level were compared under different teaching modes. Compared with the no segment method, the average accuracy of the experimental design method increased by 2.80% and 3.50%, respectively, and the action recognition rate increased by 11.50%, 8.40%, 13.60%, 13.50%, and 13.60%, respectively. Before and after the experiment, there was a significant difference in the performance of the experimental group ( P = 0.021 < 0.05 ). The results show that the traditional sports teaching mode based on video action recognition technology of big data dynamic programming algorithm can effectively improve the teaching quality of sports teaching. This research has a certain reference value to promote the current physical education teaching reform policy.

The Nondeterministic Situation Calculus

The standard situation calculus assumes that atomic actions are deterministic. But many domains involve nondeterministic actions, with problems such as fully observable nondeterministic (FOND) planning and high-level program execution requiring solutions. Various approaches have been proposed to accommodate nondeterminism on top of the standard situation calculus language, for instance by introducing nondeterministic programs as in Golog and ConGolog. But a key problem in these approaches is that they don’t clearly distinguish between choices that can be made by the agent and choices that are made by the environment, i.e., angelic vs. devilish nondeterminism. In this paper, we propose a simple extension to the standard situation calculus that accommodates nondeterministic actions and preserves Reiter’s solution to the frame problem and answering projection queries through regression. We also provide a formalization of FOND planning and show how ConGolog high-level program execution in nondeterministic domains can be defined.

Construction of Stretching-Bending Sequential Pattern to Recognize Work Cycles for Earthmoving Excavator from Long Video Sequences

Counting the number of work cycles per unit of time of earthmoving excavators is essential in order to calculate their productivity in earthmoving projects. The existing methods based on computer vision (CV) find it difficult to recognize the work cycles of earthmoving excavators effectively in long video sequences. Even the most advanced sequential pattern-based approach finds recognition difficult because it has to discern many atomic actions with a similar visual appearance. In this paper, we combine atomic actions with a similar visual appearance to build a stretching–bending sequential pattern (SBSP) containing only “Stretching” and “Bending” atomic actions. These two atomic actions are recognized using a deep learning-based single-shot detector (SSD). The intersection over union (IOU) is used to associate atomic actions to recognize the work cycle. In addition, we consider the impact of reality factors (such as driver misoperation) on work cycle recognition, which has been neglected in existing studies. We propose to use the time required to transform “Stretching” to “Bending” in the work cycle to filter out abnormal work cycles caused by driver misoperation. A case study is used to evaluate the proposed method. The results show that SBSP can effectively recognize the work cycles of earthmoving excavators in real time in long video sequences and has the ability to calculate the productivity of earthmoving excavators accurately.

NPU RGB+D Dataset and a Feature-Enhanced LSTM-DGCN Method for Action Recognition of Basketball Players

Computer vision-based action recognition of basketball players in basketball training and competition has gradually become a research hotspot. However, owing to the complex technical action, diverse background, and limb occlusion, it remains a challenging task without effective solutions or public dataset benchmarks. In this study, we defined 32 kinds of atomic actions covering most of the complex actions for basketball players and built the dataset NPU RGB+D (a large scale dataset of basketball action recognition with RGB image data and Depth data captured in Northwestern Polytechnical University) for 12 kinds of actions of 10 professional basketball players with 2169 RGB+D videos and 75 thousand frames, including RGB frame sequences, depth maps, and skeleton coordinates. Through extracting the spatial features of the distances and angles between the joint points of basketball players, we created a new feature-enhanced skeleton-based method called LSTM-DGCN for basketball player action recognition based on the deep graph convolutional network (DGCN) and long short-term memory (LSTM) methods. Many advanced action recognition methods were evaluated on our dataset and compared with our proposed method. The experimental results show that the NPU RGB+D dataset is very competitive with the current action recognition algorithms and that our LSTM-DGCN outperforms the state-of-the-art action recognition methods in various evaluation criteria on our dataset. Our action classifications and this NPU RGB+D dataset are valuable for basketball player action recognition techniques. The feature-enhanced LSTM-DGCN has a more accurate action recognition effect, which improves the motion expression ability of the skeleton data.

Embodiment and Its Influence on Informational Costs of Decision Density—Atomic Actions vs. Scripted Sequences

The importance of embodiment for effective robot performance has been postulated for a long time. Despite this, only relatively recently concrete quantitative models were put forward to characterize the advantages provided by a well-chosen embodiment. We here use one of these models, based on the concept of relevant information, to identify in a minimalistic scenario how and when embodiment affects the decision density. Concretely, we study how embodiment affects information costs when, instead of atomic actions, scripts are introduced, that is, predefined action sequences. Their inclusion can be treated as a straightforward extension of the basic action space. We will demonstrate the effect on informational decision cost of utilizing scripts vs. basic actions using a simple navigation task. Importantly, we will also employ a world with “mislabeled” actions, which we will call a “twisted” world. This is a model which had been used in an earlier study of the influence of embodiment on decision costs. It will turn out that twisted scenarios, as opposed to well-labeled (“embodied”) ones, are significantly more costly in terms of relevant information. This cost is further worsened when the agent is forced to lower the decision density by employing scripts (once a script is triggered, no decisions are taken until the script has run to its end). This adds to our understanding why well-embodied (interpreted in our model as well-labeled) agents should be preferable, in a quantifiable, objective sense.

A Comparative Study of Feature Selection Approaches for Human Activity Recognition Using Multimodal Sensory Data

Human activity recognition (HAR) aims to recognize the actions of the human body through a series of observations and environmental conditions. The analysis of human activities has drawn the attention of the research community in the last two decades due to its widespread applications, diverse nature of activities, and recording infrastructure. Lately, one of the most challenging applications in this framework is to recognize the human body actions using unobtrusive wearable motion sensors. Since the human activities of daily life (e.g., cooking, eating) comprises several repetitive and circumstantial short sequences of actions (e.g., moving arm), it is quite difficult to directly use the sensory data for recognition because the multiple sequences of the same activity data may have large diversity. However, a similarity can be observed in the temporal occurrence of the atomic actions. Therefore, this paper presents a two-level hierarchical method to recognize human activities using a set of wearable sensors. In the first step, the atomic activities are detected from the original sensory data, and their recognition scores are obtained. Secondly, the composite activities are recognized using the scores of atomic actions. We propose two different methods of feature extraction from atomic scores to recognize the composite activities, and they include handcrafted features and the features obtained using the subspace pooling technique. The proposed method is evaluated on the large publicly available CogAge dataset, which contains the instances of both atomic and composite activities. The data is recorded using three unobtrusive wearable devices: smartphone, smartwatch, and smart glasses. We also investigated the performance evaluation of different classification algorithms to recognize the composite activities. The proposed method achieved 79% and 62.8% average recognition accuracies using the handcrafted features and the features obtained using subspace pooling technique, respectively. The recognition results of the proposed technique and their comparison with the existing state-of-the-art techniques confirm its effectiveness.

Mechanized Theory of Event Structures: A Case of Parallel Register Machine

The true concurrency models, and in particular event structures, have been introduced in the 1980s as an alternative to operational interleaving semantics of concurrency, and nowadays they are regaining popularity. Event structures represent the causal dependency and conflict between the individual atomic actions of the system directly. This property leads to a more compact and concise representation of semantics. In this work-in-progress report, we present a theory of event structures mechanized in the COQ proof assistant and demonstrate how it can be applied to define certified executable semantics of a simple parallel register machine with shared memory.

Explainable Anomaly Detection in Videos Based on the Description of Atomic Actions

Anomaly detection in surveillance videos is an exhaustive and tedious task to be performed manually by humans. Many methods have been proposed to detect anomalous events by learning normal patterns and differentiate them from abnormal ones. However, these methods often suffer from false alarms, as human behaviors and environments can change over time. In addition, these methods fail to discriminate the types of anomalies that can occur, especially in anomalies performed by humans. This work presents an approach to detect anomalous events based on atomic action descriptions. It combines a tracking people method with atomic action detection and recognition network to understand video events and generate atomic descriptions. Besides detecting the anomalies, the proposed approach can also describe the anomalous action with human attributes in natural language. Anomalies are detected based on the generated descriptions of the scene. Experimental results show the effectiveness of our approach, presenting an average F1-Score of 87%.

Rank Pooling Approach for Wearable Sensor-Based ADLs Recognition

This paper addresses wearable-based recognition of Activities of Daily Living (ADLs) which are composed of several repetitive and concurrent short movements having temporal dependencies. It is improbable to directly use sensor data to recognize these long-term composite activities because two examples (data sequences) of the same ADL result in largely diverse sensory data. However, they may be similar in terms of more semantic and meaningful short-term atomic actions. Therefore, we propose a two-level hierarchical model for recognition of ADLs. Firstly, atomic activities are detected and their probabilistic scores are generated at the lower level. Secondly, we deal with the temporal transitions of atomic activities using a temporal pooling method, rank pooling. This enables us to encode the ordering of probabilistic scores for atomic activities at the higher level of our model. Rank pooling leads to a 5–13% improvement in results as compared to the other popularly used techniques. We also produce a large dataset of 61 atomic and 7 composite activities for our experiments.