Machine Learning, Classification and Situation Understanding

At the Institute of Measurement, Control and Microtechnology, methods of machine learning are used, among other things, for situational analysis, classification and situational prediction. In this way, patterns and relations are found from observations of reality using statistical learning methods.

Contact

Prof. Dr.-Ing. Klaus Dietmayer
Raum: 41.2.222
Telefon: +49 (0)731 50 27000
E-Mail | Homepage

Classification and Situation Understanding

For an intelligent and strategically reasonable driving the classification of all objects of the current vehicle environment is required. A suitable method for this task is scene labelling. An image is segmented into several object classes, in order to yield an understanding of the scene. Each pixel in the input image is assigend to an a-posteriori probability of the respective class by means of a trained neural convolutional network. The network was trained using a backpropagation algorithm with manually tagged images. Neighborhood relationships of the pixels are used to determine the class. To determine the labels, parallel algorithms, wich were implemented on a GPU, were developed at the Institute of Measurement, Control and Microtechnology.

 

Examplary result of scene labelling. The original gray scale image was pixel-wise segmented into four classes.

Situation Prediction

The prediction of the situation aims at the prediction of a future scene development. Since the future development can have a variety of manifestations, the prediction result is rather uncertain. The possible scene evolution, based on a known past, is therefore determined probabilistically.
An example of a method investigated at the institute is the Variational Bayesian Bernoulli Gaussian Mixture Model (VBBGMM). The method is able to learn the classification procedure independently with the help of training data. The gained probability model allows the integration of a priori knowledge. This enables the model to be adapted at a later stage by the integration of additional training data without a consistent storage of the previous training data.

Variational Bayesian Bernoulli Gaussian Mixture Model (VBBGMM)