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 (≈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 (≈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.

The underwater environment is characterized by hazardous conditions that make it difficult to manage and monitor even the simplest human operation. The introduction of a robot companion with the task of supporting and monitoring the divers during their activities and operations underwater can help to solve some of the problems that usually arise in this scenario. In this context, a proper communication between the diver and the robot is imperative for the success of the dive. However, the underwater environment poses a set of technical challenges which are not readily surmountable thus limiting the spectrum from which possibilities can be chosen. This paper presents the design and development of a gesture-based communication language which has been employed for the entire duration of the European project CADDY (Cognitive Autonomous Diving Buddy). This language, the Caddian, was built upon consolidated and standardized underwater gestures that are commonly used in recreational and professional diving. Its use and integration during field tests with a remotely operated underwater vehicle (ROV) is also shown.

Underwater environment is characterized by harsh conditions and is difficult to monitor. The CADDY project deals with the development of a companion robot devoted to support and to monitor human operations and activities during the dive. In this scenario the communication and correct reception of messages between the diver and the robot are essential for success of the dive goals. However, the underwater environment poses a set of technical constraints hardly limiting the communication possibilities. For such reasons the solution proposed is to develop a communication language based on the consolidated and standardized diver gestures, commonly employed during professional and recreational dives, thus leading to the definition of a CADDY language, called CADDIAN, and a communication protocol. This article focuses on the creation of the language providing alphabet, syntax and semantics: future work will explain the part of recognition of gestures that is still in progress.

This simulation research explores the informational causes of polarization and bi-polarization of opinions within groups. We define 'polarization' here as a uniform change of the opinion of the whole group in the same direction, whereas 'bi-polarization' indicates a split of two subgroups towards opposite directions. For our purposes, we have expanded the model of the Argument Communication Theory of Bi-polarization. This is an argument-based multi-agent model of opinion dynamics inspired by Persuasive Argument Theory. The original model accounts for polarization as an outcome of pure informational influence and reproduces bipolarization effects by postulating an additional mechanism of homophilous selection of communication partners. The expanded model adds two dimensions: i.e., argument strength and more sophisticated protocols of informational influence (argument communication and opinion update). Adding the first dimension, allows us to investigate whether and how the presence of stronger or weaker arguments in a discussion influences polarization and bi-polarization dynamics, as suggested by the original framework of Persuasive Arguments Theory. The second feature allows us to test whether other mechanisms related to confirmation bias and epistemic vigilance can act as a driving force of bi-polarization. For the first issue, our simulations showed that argument strength has a measurable effect. For the second, our results would indicate that, in absence of homophily, only very strong types of informational bias can lead to bi-polarization.