WARV 2014 Abstracts

Short Papers
Paper Nr: 1

Fast Incremental Objects Identification and Localization using Cross-correlation on a 6 DoF Voting Scheme


Mauro Antonello, Alberto Pretto and Emanuele Menegatti

Abstract: In this work, we propose a sparse features-based object recognition and localization system, well suited for online learning of new objects. Our method takes advantages of both depth and ego-motion information, along with salient feature descriptors information, in order to learn and recognize objects with a scalable approach. We extend the conventional probabilistic voting scheme for object the recognition task, proposing a correlation-based approach in which each object-related point feature contributes in a 6-dimensional voting space (i.e., the 6 degrees-of-freedom, DoF, object position) with a continuous probability density distribution (PDF) represented by a Mixture of Gaussian (MoG). A global PDF is then obtained adding the contribution of each feature. The object instance and pose are hence inferred exploiting an efficient mode-finding method for mixtures of Gaussian distributions. The special properties of the convolution operator for the MoG distributions, combined with the sparsity of the exploited data, provide our method with good computational efficiency and limited memory requirements, enabling real-time performances also in robots with limited resources.

Paper Nr: 2

Bio-inspired Active Vision for Obstacle Avoidance


Manuela Chessa, Saverio Murgia, Luca Nardelli, Silvio P. Sabatini and Fabio Solari

Abstract: Reliable distance estimation of objects in a visual scene is essential for any artificial vision system designed to serve as the main sensing unit on robotic platforms. This paper describes a vision-centric framework for a mobile robot which makes use of bio-inspired techniques to solve visual tasks, in particular to estimate disparity. Such framework features robustness to noise, high speed in data processing, good performance in 3D reconstruction, the possibility to orientate the cameras independently and it requires no explicit estimation of the extrinsic parameters of the cameras. These features permit navigation with obstacle avoidance allowing active exploration of the scene. Furthermore, the modular design allows the integration of new modules with more advanced functionalities.

Paper Nr: 3

Semantic Labelling of 3D Point Clouds using Spatial Object Constraints


Malgorzata Goldhoorn and Ronny Hartanto

Abstract: The capability of dealing with knowledge from the real human environment is required for autonomous systems to perform complex tasks. The robot must be able to extract the objects from the sensors’ data and give them a meaningful semantic description. In this paper a novel method for semantic labelling is presented. The method is based on the idea of connecting spatial information about the objects to their spatial relations to other entities. In this approach, probabilistic methods are used to deal with incomplete knowledge, caused by noisy sensors and occlusions. The process is divided into two stages. First, the spatial attributes of the objects are extracted and used for the object pre-classification. Second, the spatial constraints are taken into account for the semantic labelling process. Finally, we show that the use of spatial object constraints improves the recognition results.