2 results
Search Results
Now showing 1 - 2 of 2
Article Citation - WoS: 2Citation - Scopus: 3A Hybrid Approach for Semantic Image Annotation(Ieee-inst Electrical Electronics Engineers inc, 2021) Sezen, Arda; Turhan, Cigdem; Sengul, GokhanIn this study, a framework that generates natural language descriptions of images within a controlled environment is proposed. Previous work on neural networks mostly focused on choosing the right labels and/or increasing the number of related labels to depict an image. However, creating a textual description of an image is a completely different phenomenon, structurally, syntactically, and semantically. The proposed semantic image annotation framework presents a novel combination of deep learning models and aligned annotation results derived from the instances of the ontology classes to generate sentential descriptions of images. Our hybrid approach benefits from the unique combination of deep learning and semantic web technologies. We detect objects from unlabeled sports images using a deep learning model based on a residual network and a feature pyramid network, with the focal loss technique to obtain predictions with high probability. The proposed framework not only produces probabilistically labeled images, but also the contextual results obtained from a knowledge base exploiting the relationship between the objects. The framework's object detection and prediction performances are tested with two datasets where the first one includes individual instances of images containing everyday scenes of common objects and the second custom dataset contains sports images collected from the web. Moreover, a sample image set is created to obtain annotation result data by applying all framework layers. Experimental results show that the framework is effective in this controlled environment and can be used with other applications via web services within the supported sports domain.Conference Object Citation - WoS: 10Citation - Scopus: 13A Multi-Agent System Model for Partner Selection Process in Virtual Enterprise(Elsevier Science Bv, 2014) Sadigh, B. Lotfi; Arikan, F.; Ozbayoglu, A. M.; Unver, H. O.; Kilic, S. E.Virtual Enterprise (VE) is a collaboration model between multiple business partners in a value chain. VE information system deals with highly dynamic information from heterogeneous data sources. In order to manage and store dynamic VE information in the database, an ontology based VE model has been developed. To select winner enterprises in VE, a Multi Agent System (MAS) has been developed. Communication and data transition among agents and system entities are based on defined rules in VE ontology model. One of the most important contributions of agents in VE system is in partner selection step of VE formation phase. In this step several agents with different goals and strategies are collaborating and competing each other to win the negotiation procedure or maximize the profit for their assigned enterprise. Different strategies are developed for the agents depending on their appetite for winning the auction against maximizing the profit. Several simulations were run and the results are stored. These results are fed into a neural network in order to predict which enterprise will win the auction and what will be the profit margin. The motivation is to provide a forecasting agent for the customers about the outcomes of the auctions so that they can plan ahead and take the necessary action. Early results indicate such simulated multi-agent VE formations can be used in real systems. A Multi-Agent System Model for Partner Selection Process in Virtual Enterprise (C) 2014 Published by Elsevier B.V.
