Cem Karaoguz

Learn More
We present a biologically inspired architecture for incremental learning that remains resource-efficient even in the face of very high data dimensionalities (>1000) that are typically associated with perceptual problems. In particular, we investigate how a new perceptual (object) class can be added to a trained architecture without retraining, while(More)
Domestic tasks such as grasping or navigation for robotic systems can be supported by vision. However, the environment provides a vast amount of visual information and concentrating on the information related to the task being undertaken is an important job. Active vision is an approach that provides such a filtering mechanism by using camera movements to(More)
Many real-world applications in robotics have to deal with imprecisions and noise when using only a single information source for computation. Therefore making use of additional cues or sensors is often the method of choice. One examples considered in this paper is depth estimation where multiple visual and auditory cues can be combined to increase(More)
Interactions between humans or humanoids and their environment through tasks like grasping or manipulation typically require accurate depth information. The human vision system integrates various monocular and binocular depth estimation mechanisms in order to achieve robust and reliable depth perception. Such an integrated approach can be applied to(More)
In active vision systems, which direct their gaze to different visual targets in their environment, targets are represented in image coordinates and commands by which the gaze direction is changed are represented in motor coordinates. This requires knowledge of the mapping between two coordinate frames. In this work we present a robust mechanism that learns(More)
Many state of the art object classification applications require many data samples, whose collection is usually a very costly process. Performing initial model training with synthetic samples (from virtual reality tools) has been proposed as a possible solution, although the resulting classification models need to be adapted (fine-tuned) to real-world data(More)
We present a novel use for self-organizing maps (SOMs) as an essential building block for incremental learning algorithms. SOMs are very well suited for this purpose because they are inherently online learning algorithms, because their weight updates are localized around the best-matching unit, which inherently protects them against catastrophic forgetting,(More)
  • 1