We use proprietary and third party's cookies to improve your experience and our services, identifying your Internet Browsing preferences on our website; develop analytic activities and display advertising based on your preferences. If you keep browsing, you accept its use. You can get more information on our Cookie Policy
Cookies Policy
SAFECITY.Story.DATA/CONTEXT MANAGEMENT.Data.MultimediaAnalysis.ModelRetraining - FIWARE Forge Wiki

SAFECITY.Story.DATA/CONTEXT MANAGEMENT.Data.MultimediaAnalysis.ModelRetraining

From FIWARE Forge Wiki

Jump to: navigation, search
Name ModelRetraining
Goal To enhance the detection efficiency of the video analytics application
Description It is often the case that learned models in classification algorithms or clusters in clustering algorithms, due to insufficiency of training data or noisy data, are often not optimal or fail to represent the full potential of a specific concept or entity. With the provision of new video data through the constant monitoring process, these serve as input for model retraining (running the classification and clustering algorithms to reproduce models or clusters) in case that this process leads to higher accuracy of the latter. Furthermore, the validation of detection results by human operators at the Command Centre facilitates this process even more and with a higher accuracy. Both these tasks are considered in a component dealing with model retraining.
Version 1.0 Source Athens Information Technology - AIT, Sofia Tsekeridou
Scope Platform Epic {{{Epic}}} Theme
Status Pending MoSCoW priority MST Relative priority
Chapter Data/Context Management, Enabler Multimedia Analysis
Stakeholder Owner Athens Information Technology - AIT, Sofia Tsekeridou
Complexity Rationale

Personal tools
Create a book