A5047731654- Robot Manipulation and Learning
- Hand Gesture Recognition Systems
- Human Pose and Action Recognition
- Robotics and Sensor-Based Localization
- Soft Robotics and Applications
- Industrial Vision Systems and Defect Detection
- Advanced Vision and Imaging
- Energy Efficient Wireless Sensor Networks
- Teleoperation and Haptic Systems
- Image and Object Detection Techniques
- Indoor and Outdoor Localization Technologies
- Energy Harvesting in Wireless Networks
- Mobile Ad Hoc Networks
- 3D Surveying and Cultural Heritage
- Reinforcement Learning in Robotics
- Advanced Neural Network Applications
- IoT-based Smart Home Systems
- Adversarial Robustness in Machine Learning
FPT University
2023-2024
Institut de Recherche en Informatique et Systèmes Aléatoires
2014-2016
Université de Rennes
2016
Institut national de recherche en informatique et en automatique
2014
Université Rennes 2
2014
Accurate detection and estimation of pallet poses from color depth data (RGB-D) are integral components many in advanced warehouse intelligent systems. State-of-the art object pose methods follow a two-stage process, relying on off-the-shelf segmentation or the initial stage subsequently predicting objects using cropped images. The patches may include both target irrelevant information, such as background other objects, leading to challenges handling pallets settings with heavy occlusions...
Object recognition and pose estimation are critical components in autonomous robot manipulation systems, playing a crucial role enabling robots to interact effectively with the environment. During actual execution, must recognize object current scene, estimate its pose, then select feasible grasp from pre-defined configurations. While most existing methods primarily focus on estimation, they often neglect graspability reachability aspects. This oversight can lead inefficiencies failures...
Hand-object configuration recovery is an important task in computer vision. The estimation of pose and shape for both hands objects during interactive scenarios has various applications, particularly augmented reality, virtual or imitation-based robot learning. problem challenging when the hand interacting with environment, as this setting features extreme occlusions non-trivial deformations. While existing works treat estimating configurations (that parameters) isolation from parameters...
The task of object pose estimation in computer vision heavily relies on both color (RGB) and depth (D) images to provide crucial appearance geometric information, assisting algorithms understanding occlusions geometry, thereby enhancing accuracy. However, the dependency specialized sensors capable capturing poses challenges terms cost availability. Consequently, researchers are exploring methods estimate solely from RGB images. Nevertheless, this approach encounters difficulties handling...
Grasp detection plays a pivotal role in robotic manipulation, allowing robots to interact with and manipulate objects their surroundings. Traditionally, this has relied on three-dimensional (3D) point cloud data acquired from specialized depth cameras. However, the limited availability of such sensors real-world scenarios poses significant challenge. In many practical applications, operate diverse environments where obtaining high-quality 3D may be impractical or impossible. This paper...
In recent years, many MAC protocols for wireless sensor networks have been proposed and most of them are evaluated using Matlab simulator and/or network simulators (OMNeT++, NS2, etc). However, a static behavior few simulations available adaptive protocols. Specially, in OMNeT++/MiXiM, there energy-efficient WSNs (B-MAC & L-MAC) no ones. To this end, the TAD-MAC (Traffic Aware Dynamic MAC) protocol has simulated OMNeT++ with MiXiM framework implementation details given paper. The...