Explanatory feedback to user generated media


User generated media contributions are of growing importance in professional media production, e.g., for emerging events where no news capture teams is present yet, or for large-scale events such as music festivals, where many things happen in parallel. For example, radio and TV stations frequently ask users for contributing images and videos for specific topics. However, not all of them are usable, partly due to the technical quality or because they do not show the subject of interest. Manually selecting content is not feasible, but the task can be supported by automatic content and video analysis. The reaming challenge is to provide a meaningful response to the contributor.


Available content analysis algorithms can be used to extract information about the types of objects present, or the location. Similarly, automatic quality analysis tools allow assessing properties such as image noise, over-/underexposure or shakiness of video. All these algorithms provide fine grained measurements about the content. The aim is to provide feedback to the contributor, which helps them to understand why a specific image or video may not be suitable in terms of quality or content, and provide a response that is explanatory and encourages improvement.

This is expected to be achieved by developing a pipeline of the following steps:

  • Based on a sematic representation of the topic, select the content properties that are desired and unwanted.
  • Run a set of existing content and quality analysis tools, develop an aggregation strategy of the results.
  • Match the analysis results against the model of the expected content and quality and derive a score for the item.
  • If the score is low, identify the aspects that contributed to decreasing it.
  • Generate a response to the user in a semantic representation, which allows straight forwardly deriving textual responses in different languages.

Required Skills

  • Knowledge of semantic web technologies
  • Basic understanding of image analysis, image/video processing
  • Software development: C ++, Java


As soon as possible


This master thesis is scheduled for approx. 6 months.


JOANNEUM RESEARCH offers interested students cooperation with all Austrian but also European or non-European educational institutions and is willing to assist you looking for a supervisor, if necessary.


We are looking forward to receiving convincing applications only via e-mail to:

JOANNEUM RESEARCH Forschungsgesellschaft mbH
DIGITAL – Institute for Information and Communication Technologies
Subject: DIG Mastertheses – „Thema“

For content questions
Phone:+43 316 876-1218
Fax+43 316 8769-1191