INSANE Dataset – Cross-Domain UAV Data Sets with Increased Number of Sensors for developing Advanced and Novel Estimators
A multi-sensor cross-domain UAV data set (18 sensors) with accurate and absolute 6 DoF ground truth. The scenarios include indoor flights in a controlled environment with motion capture ground truth, outdoor-to-indoor transition flights with continuous ground truth, and extensive coverage of Mars analog data with the same vehicle. Mars analog data includes segments with various ground structures, cliff flight over, and cliff-wall traversing trajectories for mapping.
This data set is ideal for testing novel algorithms with real-world sensor data and corresponding effects such as sensor degradation. Dedicated raw data for customized sensor calibration routines and vibration data for vehicle integrity tests are provided.
Metric Volume Estimation of Fruits and Vegetables
A challenging data set for metric volume estimation of fruits and vegetables. The data set consists of monocular video sequences recorded with a handheld smartphone while moving around different types of fruits and vegetables. It contains the video frames and recorded inertial data during the videos as well as ground-truth volume information. It has been constructed to mirror real-world use cases incorporating varying backgrounds, different angles and distances from the object of interest.
PoET: Pose Estimation Transformer for Single-View, Multi-Object 6D Pose Estimation
PoET is a transformer-based framework that takes a single RGB-image as input to simultaneously estimate the 6D pose, namely translation and rotation, for every object present in the image. It takes the detections and feature maps of an object detector backbone and feeds this additional information into an attention-based transformer. Our framework can be trained on top of any object detector framework. Any additional information that is not contained in the raw RGB image, e.g. depth maps or 3D models, is not required. We achieve state-of-the-art-results on challenging 6D object pose estimation datasets. Moreover, PoET can be utilized as a pose sensor in 6D localization tasks.
AIVIO: Closed-loop, Object-relative Navigation of UAVs with AI-aided Visual Inertial Odometry
Object-relative mobile robot navigation is essential for a variety of tasks, e.g. autonomous critical infrastructure inspection, but requires the capability to extract semantic information about the objects of interest from raw sensory data. While deep learning-based (DL) methods excel at inferring semantic object information from images, such as class and relative 6 degree of freedom (6-DoF) pose, they are computationally demanding and thus often not suitable for payload constrained mobile robots. In this letter we present a real-time capable unmanned aerial vehicle (UAV) system for object-relative, closed-loop navigation with a minimal sensor configuration consisting of an inertial measurement unit (IMU) and RGB camera. Utilizing a DL-based object pose estimator, solely trained on synthetic data and optimized for companion board deployment, the object-relative pose measurements are fused with the IMU data to perform object-relative localization. We conduct multiple real-world experiments to validate the performance of our system for the challenging use case of power pole inspection. An example closed-loop flight is presented in the supplementary video.
