Tags:Data-driven approach, Deep Learning, Knowledge-driven approach, Ontology and Semantic gap
Abstract:
Nowadays, imaging and artificial vision are becoming important fields of research. Nevertheless, despite the high accuracy that has been achieved by their techniques, most of them remain “quantitative-approaches”; their efficiency depends on the computational capacity and requires a large amount of data for training and testing. In addition, there are still many deficiencies such as the semantic gap between the low-level visual information and high-level semantic knowledge, the recognition of complex scenes and so on. With the aim of providing less costly and more efficient solutions, the new trend is to make it possible for the aforementioned fields to support knowledge representation techniques i.e. “qualitative-approaches”, which is at first sight ontological. This article reviews several research-works that have demonstrated the effectiveness of combining those approaches. This combination has certain advantages, among which: closing its semantic gap, reducing the data rate required by making logical inferences, enhancing the querying capability.
Integrating Ontology with Imaging and Artificial Vision for a High-Level Semantic; a Review