Dinh-Cuong Hoang
Befattning: Forskare Organisation: Institutionen för naturvetenskap och teknikE-post: cuong.hoang@oru.se
Telefon: 019 301359
Rum: T1224
Om Dinh-Cuong Hoang
Project title and summary in brief:
Object recognition in 3D models reconstructed from RGB-D sensor data.
The goal of object recognition and localization is to find the position and orientation of a set of objects of interest in 3-D space. This task still poses a significant challenge, because the target objects can be any forms, shapes, and have six degrees of freedom (DOF). In addition, the targets may be arranged in a challenging manner in the scene. Because there are many objects with same shape and appearance, it is not easy to recognize the right object and place targets at the defined position, even for a human. The main goals of this study are robust recognition and localization of a set of target objects, using RGB-D sensor data as input and evaluating the suitability of the detected poses to an automated picking system.
Publikationer
Artiklar i tidskrifter
- Hoang, D. , Lilienthal, A. & Stoyanov, T. (2020). Object-RPE: Dense 3D Reconstruction and Pose Estimation with Convolutional Neural Networks. Robotics and Autonomous Systems, 133.
- Hoang, D. , Lilienthal, A. & Stoyanov, T. (2020). Panoptic 3D Mapping and Object Pose Estimation Using Adaptively Weighted Semantic Information. IEEE Robotics and Automation Letters, 5 (2), 1962-1969.
Doktorsavhandlingar
- Hoang, D. (2021). Vision-based Perception For Autonomous Robotic Manipulation. (Doctoral dissertation). Örebro: Örebro University.
Konferensbidrag
- Hoang, D. , Stork, J. A. & Stoyanov, T. (2022). Context-Aware Grasp Generation in Cluttered Scenes. I: 2022 International Conference on Robotics and Automation (ICRA). Konferensbidrag vid IEEE International Conference on Robotics and Automation (ICRA 2022), Philadelphia, USA, May 23-27, 2022. (ss. 1492-1498). IEEE.
- Hoang, D. , Stoyanov, T. & Lilienthal, A. J. (2019). Object-RPE: Dense 3D Reconstruction and Pose Estimation with Convolutional Neural Networks for Warehouse Robots. I: 2019 European Conference on Mobile Robots, ECMR 2019 Proceedings. Konferensbidrag vid 2019 European Conference on Mobile Robots (ECMR), Prague, Czech Republic, September 4-6, 2019. IEEE.