Markus Thom

Omnidirectional Cameras as Backing-Up Aid

Omnidirectional cameras are well suited for maneuvering tasks due to to their large field of view. We mounted two catadioptric cameras on a vehicle and provide the driver with a birds-eye view of the surrounding area behind the vehicle as well as the area on the left and right hand side. In order to enlarge the field of view of the birds-eye view image, we will show an extension to the birds-eye view image that provides a larger field of view and thus simplifies challenging maneuvering tasks. An overlay onto the resulting image shows the motion path of the vehicle while reversing. This motion path is connected to the steering angle. The driver can easily see where the vehicle will move when the steering angle is not changed throughout the movement.

Supervised Matrix Factorization with sparseness constraints and fast inference

Non-negative Matrix Factorization is a technique for decomposing large data sets into bases and code words, where all entries of the occurring matrices are non-negative. A recently proposed technique also incorporates sparseness constraints, in such a way that the amount of nonzero entries in both bases and code words becomes controllable. This paper extends the Non-negative Matrix Factorization with Sparseness Constraints. First, a modification of the optimization criteria ensures fast inference of the code words. Thus, the approach is real-time capable for use in time critical applications. Second, in case a teacher signal is associated with the samples, it is considered in order to ensure that inferred code words of different classes can be well distinguished. Thus, the derived bases generate discriminative code words, which is a crucial prerequisite for training powerful classifiers. Experiments on natural image patches show, similar to recent results in the field of sparse coding algorithms, that Gabor-like filters are minimizing the reconstruction error while retaining inference capabilities. However, applying the approach with incorporation of the teacher signal to handwritten digits yields morphologically completely different bases, while achieving superior classification results.

Training of sparsely connected MLPs

Sparsely connected Multi-Layer Perceptrons (MLPs) differ from conventional MLPs in that only a small fraction of entries in their weight matrices are nonzero. Using sparse matrix-vector multiplication algorithms reduces the computational complexity of classification. Training of sparsely connected MLPs is achieved in two consecutive stages. In the first stage, initial values for the networks parameters are given by the solution to an unsupervised matrix factorization problem, minimizing the reconstruction error. In the second stage, a modified version of the supervised backpropagation algorithm optimizes the MLPs parameters with respect to the classification error. Experiments on the MNIST database of handwritten digits show that the proposed approach achieves equal classification performance compared to a densely connected MLP while speeding-up classification by a factor of seven.