CADDY Underwater Stereo-Vision Dataset for Human-Robot Interaction (HRI) in the Context of Diver Activities

In this article, we present a novel underwater dataset collected from several field trials within the EU FP7 project Cognitive autonomous diving buddy (CADDY), where an Autonomous Underwater Vehicle (AUV) was used to interact with divers and monitor their activities. To our knowledge, this is one of the first efforts to collect a large public dataset in underwater environments with the purpose of studying and boosting object classification, segmentation and human pose estimation tasks. The first part of the dataset contains stereo camera recordings (approximate to 10 K) of divers performing hand gestures to communicate with an AUV in different environmental conditions. The gestures can be used to test the robustness of visual detection and classification algorithms in underwater conditions, e.g., under color attenuation and light backscatter. The second part includes stereo footage (approximate to 12.7 K) of divers free-swimming in front of the AUV, along with synchronized measurements from Inertial Measurement Units (IMU) located throughout the diver's suit (DiverNet), which serve as ground-truth for human pose and tracking methods. In both cases, these rectified images allow the investigation of 3D representation and reasoning pipelines from low-texture targets commonly present in underwater scenarios. This work describes the recording platform, sensor calibration procedure plus the data format and the software utilities provided to use the dataset.