Developing a Context-Aware System for Providing Intelligent Robot Services

Personalized context-aware systems for sustainable agriculture development using ubiquitous devices and adaptive learning

Robots are typically equipped with limited computing resources and robot services need to different computing requirement such as sensors, actuators, computational capabilities and time constraint of service in network-based robots. In this paper, we propose a dynamic management scheme of service contexts for network-based robots to execute various services efficiently in ubiquitous environment. By configuring the service contexts for each service, the proposed method efficiently uses the limited computing resources of a robot and improves the overall performance of a network-based robot system. The proposed method is experimented on the real network-based robot system. The results show that the time constraint for services is satisfied and the resource utilization is improved. It can provide the extendibility and flexibility for services by aptly modifying service contexts.

Most context-aware recommender systems in the literature that use context modelling have the tendency to develop domain and application specific context models that limit, even eliminate any reuse and sharing capabilities. Developers and researchers in the field struggle to design their own context models without having a good understanding of context and without using any reference models for guidance, often resulting in overspecialized, inefficient or incomplete context models. In this work we build upon prior work to propose an enhanced online context modelling system for Context-Aware Recommender Systems. The system supports CARS developers in the process of building their own context models from scratch, while it supports at the same time sharing and reuse of the models among developers. The system was tested with a real dataset with positive results, as it was able to support context model development with instructions to the developer, model comparison, useful statistics, recommendations of similar models, as well as alternative views of context models to aid the developer’s task.

Recent developments of information technologies are leading the advent of the era of ubiquitous healthcare service, which means healthcare services at any time and at any places. In a home, the ubiquitous healthcare service requires some context aware system for smart services using home network. In this paper, we propose a systematic design tool of a context aware system for ubiquitous healthcare service in a smart home, which is composed of a home gateway, a home server, a PDA, home network healthcare devices, and wearable sensors. We developed a context aware framework to send information from wearable sensors or healthcare devices to healthcare service entities as a middleware to solve the interoperability problem among sensor makers and healthcare service providers. And, we propose a design tool of ubiquitous healthcare services with the context aware framework in a systematic way. In order to show the feasibility of the proposed system, an application example is given, which is applied to a self check service.

Adapting intelligent context-aware systems (CAS) to future operating rooms (OR) aims to improve situational awareness and provide surgical decision support systems to medical teams. CAS analyzes data streams from available devices during surgery and communicates real-time knowledge to clinicians. Indeed, recent advances in computer vision and machine learning, particularly deep learning, paved the way for extensive research to develop CAS. In this work, a deep learning approach for analyzing laparoscopic videos for surgical phase recognition, tool classification, and weakly-supervised tool localization in laparoscopic videos was proposed. The ResNet-50 convolutional neural network (CNN) architecture was adapted by adding attention modules and fusing features from multiple stages to generate better-focused, generalized, and well-representative features. Then, a multi-map convolutional layer followed by tool-wise and spatial pooling operations was utilized to perform tool localization and generate tool presence confidences. Finally, the long short-term memory (LSTM) network was employed to model temporal information and perform tool classification and phase recognition. The proposed approach was evaluated on the Cholec80 dataset. The experimental results (i.e., 88.5% and 89.0% mean precision and recall for phase recognition, respectively, 95.6% mean average precision for tool presence detection, and a 70.1% F1-score for tool localization) demonstrated the ability of the model to learn discriminative features for all tasks. The performances revealed the importance of integrating attention modules and multi-stage feature fusion for more robust and precise detection of surgical phases and tools.

Humans reliably categorize configurations of facial actions into specific emotion categories, leading some to argue that this process is invariant between individuals and cultures. However, growing behavioral evidence suggests that factors such as emotion-concept knowledge may shape the way emotions are visually perceived, leading to variability-rather than universality-in facial-emotion perception. Understanding variability in emotion perception is only emerging, and the neural basis of any impact from the structure of emotion-concept knowledge remains unknown. In a neuroimaging study, we used a representational similarity analysis (RSA) approach to measure the correspondence between the conceptual, perceptual, and neural representational structures of the six emotion categories Anger, Disgust, Fear, Happiness, Sadness, and Surprise. We found that subjects exhibited individual differences in their conceptual structure of emotions, which predicted their own unique perceptual structure. When viewing faces, the representational structure of multivoxel patterns in the right fusiform gyrus was significantly predicted by a subject's unique conceptual structure, even when controlling for potential physical similarity in the faces themselves. Finally, cross-cultural differences in emotion perception were also observed, which could be explained by individual differences in conceptual structure. Our results suggest that the representational structure of emotion expressions in visual face-processing regions may be shaped by idiosyncratic conceptual understanding of emotion categories.

Context awareness and context modeling have raised their call for attention and importance quite recently in the world of contact centers for better customer services and relations. To address the issues of contextual reuse and contextual inconsistencies, there is a need for context models that foster context reuse and support ease of retrieving right kind of information by providing appropriate abstractions of contextual information. In this paper, we propose a hybrid context model based on multilevel situation theory and ontology. Multilevel situation theory is based on an intermediate level micro situation that is used to create entity specific situations. Such micro situations are used to enhance a situation so as to enable consistent contextual reasoning. The context model is hybrid in the sense that the proposed multilevel situation theory is defined based on the entities, entity attributes, and entity relationships as described in the contact center ontology. We also present an architectural overview of context aware processing system to build a rapid prototype that demonstrates the hybrid context model.

The rapid advancement of Internet of Things (IoT) technology has opened new frontiers in inventory management by enabling real-time, data-driven decision-making. This paper presents a comprehensive framework for an IoT-enabled intelligent inventory management system (IIMS) that integrates weight sensors, Radio Frequency Identification (RFID), computer vision, and neurofuzzy logic for optimizing inventory control. The proposed system aims to enhance inventory monitoring accuracy, automate item identification and placement verification, and optimize reorder levels. Weight sensors track inventory depletion, while RFID tags enable fast and precise identification of items. Computer vision is used to detect correct and incorrect placement of items, ensuring accuracy in warehouse operations. A neuro-fuzzy model is implemented to dynamically adjust reorder points based on inventory levels and demand fluctuations. This framework facilitates seamless, real-time inventory tracking and decision-making, improving operational efficiency, reducing human errors, and supporting timely replenishment decisions. The study underscores the system's potential to transform inventory management practices by integrating IoT technologies, making it highly adaptable to various industries.

This paper presents an approach of user interface adaptation based on context awareness system and context modeling in order to ensure accessibility for disabled people in pervasive environment. The challenge is how to design and generate accessible software systems that accommodate dynamically to changing context.

I think that the main reason why we do not understand the general principles ofhow cognition works, is not the excessive complexity of cognitive phenomena, but the lack of the conceptual and methodological tools to properly address the problem. In this paper I show three principles that could be proposed asfoundation of cognitive processes. I also briefly show some points on the toolsthat I think it is necessary to develop for this purpose

Knowledge representation with SOUL

In this paper, we describe practical implementation of a framework for development and integration of network-based humanoid robot system. The characteristics of the framework are dynamically to connect, execute and disconnect humanoid robot and service components related to the robot, and to provide simplification process about development of service components and their integration task. In addition, using an error-handling mechanism, the framework is able to recovery the components interrupted by the unexpected error. Also, the stability of the network framework is guaranteed through verification process of developed service components.

This paper presents Spirit, a lightweight programming framework for intelligent system. Ease of use and tiny development effort are two key features considered during the design. Spirit is mainly a template class, which provides a configure table, commonly used libraries in system control and computer vision, interprocess communication for parallel programming, hardware middleware for different hardware products, and customizable algorithm functions. Developers utilize Spirit to create modules or even systems for their intelligent applications very quickly as well as using powerful tools in Spirit for convenient debugging. We compared Spirit with existing robot frameworks and proposed a new design criterion. Results show that Spirit gives a better experience on rapid system design and debugging. We used it when developing an intelligent vehicle system.

A common view in the reading literature is that words are often recognized by recourse to phonological recoding. The question addressed here concerns whether phonological recoding can be controlled, or whether it is computed (a) even when it is completely irrelevant to the task, and (b) can only hurt performance. Two Stroop type experiments are reported in which participants used a key press to indicate the print colour of an (irrelevant) letter string that sounded like a colour (e.g., BLOO). The sound of the colour word was always incongruent with the required response. Colour identification performance was impaired by the irrelevant letter strings relative to two baseline conditions. These results and others are consistent with the conclusion that a phonological code is computed under a variety of circumstances. The hypothesis that skilled readers transform print on the page into phonology in order to recognize words has been intensively pursued over the last several decades. Many investigators favor a single route model in which phonological coding is a necessary preliminary to visual recognition (e.g., Carello, Turvey, & Lukatela, 1992; Frost, 1998; Lukatela & Turvey, 1991, 1994; Perfetti, Zhang, & Berent, 1992; Pexman & Lupker, 1995; Van Orden et al. 1992; Van Orden, Pennington & Stone, 1990). Other investigators argue in favor of the co-existence of a reading routine that activates lexical/semantic knowledge without prior phonological computation. This orthographic route operates in parallel with the phonological route (e.g., Baluch & Besner, 1991; Besner & Smith, 1992; Buchanan & Besner, 1993; Carr & Pollatsek, 1985; Coltheart, Curtis, Atkins, & Haller, 1993; Coltheart & Rastle, 1994; Davelaar, Coltheart, Jonasson & Besner, 1978; Herdman & Beckett, 1996; Paap, Noel & Johansen 1992; Plaut, McClelland, Seidenberg & Patterson, 1996; Waters & Seidenberg, 1985). How Controllable Is Phonological Computation? One consequence of the persistent report of phonological effects in visual recognition (e.g., spelling-sound regularity, homophone, and pseudohomophone effects) is a wide spread interest in a related but distinct issue that is the focus of the present investigation. That is, to what extent can the computation of a phonological code be thought of as a controllable set of processes given a printed input, as distinct from the issue of whether it is possible to exert any control over whether the products of phonological coding are used in making decisions about a response? Clearly, failing to find an effect of a phonological manipulation in some task speaks more directly to the latter issue than it does to the former because it might not be possible to control whether a phonological code is computed, but possible to control whether the products of such coding are used in formulating a response. In that vein, some investigators emphasize that there are stimulus and contextual manipulations which encourage participants to rely on the products of a nonphonological pathway in tasks like perceptual identification, naming, lexical decision, and semantic categorization (e.g., Baluch & Besner, 1991; Besner, Twilley, McCann, & Seergobin, 1990; Davelaar, et al., 1978; Hawkins et al., 1976; Hudson & Bergman, 1985; Jared & Seidenberg, 1991; Paap et al., 1992; Seidenberg, Peterson, MacDonald, & Plaut, 1996; Tabossi & Laghi, 1992; Verstaen, Humphreys, Olson, & D'Ydewalle, 1996; Waters & Seidenberg, 1985). Other investigators take a stronger line and argue that the computation of phonology is obligatory, and even automatic (among many others, see Carello, Turvey, & Lukatela, 1992; Frost, 1997; Glushko, 1979; Katz & Frost, 1992; Lukatela & Turvey, 1994; Perfetti et al., 1992; Van Orden et al., 1992). For example: ...effect of phonological primes suggests a leading role in visual perception for a fast acting, automatic, assembled phonology. …

Segmentation and classification of objects in images is one of the most important and yet one of the most complex problems in computer vision. Many different things contribute to the complexity of this problem, but amongst those that are most significant are variations in object appearance due to various perspectives, changes in scene illumination and typically a large number of partially occluded objects in a scene. Humans have the ability to successfully deal with these complications and to recognize objects in scenes without seemingly much effort. However, to aid their recognition process, humans use different sources of information such as previous experiences, context of objects placed in a scene and rules learned in the past that suggest how the physical world should appear and behave. Computer vision is trying to replicate the same strategies used by human vision and apply them to computer systems with the aim of improving typical tasks of detection, localization and classification of image objects. In this work we propose a new model for natural image object classification using contextual information at the level of image segments. Context modeling is largely independent of appearance-based classification and proposed model enables simple upgrade of existing systems with information from global and/or local context. Context modeling is based on non-parametric use of appearance-based classification results which is a novel approach compared to previous systems that model context on a limited number of rules expressed with a fixed set of parameters. Model implementation resulted in a system that, in our simulations, showed stable improvement of the appearance-based object classification. DOI: http://dx.doi.org/10.5755/j01.itc.46.1.13610