ENNS Best Paper Award

ENNS Best Paper Award
Zdenek Straka and Matej Hoffmann were awarded ENNS Best Paper Award at 26th International Conference on Artificial Neural Networks (ICANN17) for their paper Learning a Peripersonal Space Representation as a Visual-Tactile Prediction Task.

Learning for Active 3D Mapping

Učící se aktivní 3D mapování.
Paper of authors K. Zimmermann, T. Petricek, V. Salansky, and T. Svoboda was accepted as oral presentation at ICCV 2017! We propose an active 3D mapping method for depth sensors, which allow individual control of depth-measuring rays, such as the newly emerging solid-state lidars. The method simultaneously (i) learns to reconstruct a dense 3D occupancy …read more

Non-rigid object detection with local interleaved sequential alignment

Detekce nerigidních objektů
The successively evaluated features used in a sliding window detection process to decide about object presence/absence also contain knowledge about object deformation. We exploit these detection features to estimate the object deformation. Estimated deformation is then immediately applied to not yet evaluated features to align them with the observed image data. In our approach, the …read more

Data fusion for localization and mapping

Data fusion for localization and mapping
We design and evaluate a data fusion system for localization of a mobile skid-steer robot intended for USAR missions. We exploit a rich sensor suite including both proprioceptive (inertial measurement unit and tracks odometry) and exteroceptive sensors (omnidirectional camera and rotating laser rangefinder). To cope with the specificities of each sensing modality (such as significantly …read more

Adaptive traversability

Adaptive traversability
Adaptive traversability we understand by means of autonomous motion control adapting the robot morphology—configuration of articulated parts and their compliance—to traverse unknown complex terrain with obstacles in an optimal way. The robot measures its state (like: orientation angles, flipper mode, …) and the terrain (digital elevation model). We learn the optimal policy from loosely annotated …read more

Industrial robotics

Průmyslová robotika
The Center for Machine Perception has an ongoing interest in industrial robots with the aim to broaden their capabilities. An example of such an endeavour is the European project CloPeMa. See the CloPeMa demos! The project was finished in 2015 with an excellent evaluation.

Multi-robot Systems

Multirobotické systémy
The research group of Multi-Robot Systems studies issues related to motion planning, control and coordination of teams of ground, aerial and modular robots. See the MRS research group demo page for details!

Application of Projective Reconstruction Based on Cake Configuration

Application of Projective Reconstruction Based on Cake Configuration
Martin Urban, Tomas Pajdla, Tomas Werner, Vaclav Hlavac Center for Machine Perception Czech Technical University, Prague urbanm@cmp.felk.cvut.cz,  Reconstruction was done in the following steps: Corresponding points were marked by mouse in 10 sequence images. Polygons were marked, each in one image. Projective reconstruction via trifocal tensors in CAKE configurations was done (See Tech. report Urban-TR7-99.ps.gz) …read more

CMP Structure-from-Motion Web Service

Webová služba CMP
The team composed of Tomáš Pajdla, Michal Havlena, Akihiko Torii, Michal Jančošek, Zuzana Kúkelová and Jan Heller published a CMP SfM Web Service which provides a remote access to the 3D reconstruction systems developed in Center for Machine Perception. See the gallery for examples.

Air Traffic Control

Řízení letového provozu
Air Traffic Control scenario provides an open multi-agent test bed that is used for air traffic simulation. ATC is build on the A-globe platform using Aglobex simulation framework. The project aim is to deploy the multi-agent technology for agent based aircraft deconfliction (collision avoidance) among several autonomous aerial vehicles (manned as well as unmanned).
Responsible person: Petr Pošík