Semi-Automatic Semantic Annotation of Images

Detailed, consistent semantic annotation of large collections of multimedia data is difficult and time- consuming. In domains such as eScience, digital curation and industrial monitoring, fine-grained high- quality labeling of regions enables advanced semantic querying, analysis and aggregation and supports collaborative research. Manual annotation is inefficient and too subjective to be a viable solution. Automatic solutions are often highly domain or application specific, require large volumes of annotated training corpi and, if using a `black box´ approach, add little to the overall scientific knowledge. This article evaluates the use of simple artificial neural networks to semantically annotate micrographs and discusses the generic process chain necessary for semi-automatic semantic annotation of images.