Online Semi-Supervised Classification/Clustering:
- Is obtaining labelled training data costly?
- Does the distribution of your data change over time?
- Are you dealing with data streams?
- Are you confronted with partially labelled data ?
If yes, don't hesitate to look at our research work which illustrates the application of a novel online semi-supervised model with application in online video segmentation. Long story short:
Given a few user-labeled data points the initial model is learned and then the class membership of the remaining data points in the current and subsequent time instants are estimated and propagated in an on-line fashion. Furthermore, the tracking capabilities of the Kalman filter is used to provide the labels of the objects in motion and thus regularizing the solution obtained by the MSS-KSC algorithm.