Enabling robots to understand human gaze target is a crucial step to allow capabilities in downstream tasks, for example, attention estimation and movement anticipation in real-world human-robot interactions. Prior works have addressed the in-frame target localization problem with data-driven approaches by carefully removing out-of-frame samples. Vision-based gaze estimation methods, such as OpenFace, do not effectively absorb background information in images and cannot predict gaze target in situations where subjects look away from the camera. In this work, we propose a system to address the problem of 360-degree gaze target estimation from an image in generalized visual scenes. The system, named GazeTarget360, integrates conditional inference engines of an eye-contact detector, a pre-trained vision encoder, and a multi-scale-fusion decoder. Cross validation results show that GazeTarget360 can produce accurate and reliable gaze target predictions in unseen scenarios. This makes a first-of-its-kind system to predict gaze targets from realistic camera footage which is highly efficient and deployable. Our source code is made publicly available at: https://github.com/zdai257/DisengageNet.

A major challenge to deploying robots widely is navigation in human-populated environments, commonly referred to as social robot navigation . While the field of social navigation has advanced tremendously in recent years, the fair evaluation of algorithms that tackle social navigation remains hard because it involves not just robotic agents moving in static environments but also dynamic human agents and their perceptions of the appropriateness of robot behavior. In contrast, clear, repeatable, and accessible benchmarks have accelerated progress in fields like computer vision, natural language processing and traditional robot navigation by enabling researchers to fairly compare algorithms, revealing limitations of existing solutions and illuminating promising new directions. We believe the same approach can benefit social navigation. In this article, we pave the road toward common, widely accessible, and repeatable benchmarking criteria to evaluate social robot navigation. Our contributions include (a) a definition of a socially navigating robot as one that respects the principles of safety, comfort, legibility, politeness, social competency, agent understanding, proactivity, and responsiveness to context, (b) guidelines for the use of metrics, development of scenarios, benchmarks, datasets, and simulators to evaluate social navigation, and (c) a design of a social navigation metrics framework to make it easier to compare results from different simulators, robots, and datasets.

Abstract Its numerous applications make multi-human 3D pose estimation a remarkably impactful area of research. Nevertheless, it presents several challenges, especially when approached using multiple views and regular RGB cameras as the only input. First, each person must be uniquely identified in the different views. Secondly, it must be robust to noise, partial occlusions, and views where a person may not be detected. Thirdly, many pose estimation approaches rely on environment-specific annotated datasets that are frequently prohibitively expensive and/or require specialised hardware. Specifically, this is the first multi-camera, multi-person data-driven approach that does not require an annotated dataset. In this work, we address these three challenges with the help of self-supervised learning. In particular, we present a three-staged pipeline and a rigorous evaluation providing evidence that our approach performs faster than other state-of-the-art algorithms, with comparable accuracy, and most importantly, does not require annotated datasets. The pipeline is composed of a 2D skeleton detection step, followed by a Graph Neural Network to estimate cross-view correspondences of the people in the scenario, and a Multi-Layer Perceptron that transforms the 2D information into 3D pose estimations. Our proposal comprises the last two steps, and it is compatible with any 2D skeleton detector as input. These two models are trained in a self-supervised manner, thus avoiding the need for datasets annotated with 3D ground-truth poses.

Current, self-supervised depth estimation architectures rely on clear and sunny weather scenes to train deep neural networks. However, in many locations, this assumption is too strong. For example in the UK (2021), 149 days consisted of rain. For these architectures to be effective in real-world applications, we must create models that can generalise to all weather conditions, times of the day and image qualities. Using a combination of computer graphics and generative models, one can augment existing sunny-weather data in a variety of ways that simulate adverse weather effects. While it is tempting to use such data augmentations for self-supervised depth, in the past this was shown to degrade performance instead of improving it. In this paper, we put forward a method that uses augmentations to remedy this problem. By exploiting the correspondence between unaugmented and augmented data we introduce a pseudo-supervised loss for both depth and pose estimation. This brings back some of the benefits of supervised learning while still not requiring any labels. We also make a series of practical recommendations which collectively offer a reliable, efficient framework for weather-related augmentation of self-supervised depth from monocular video. We present extensive testing to show that our method, Robust-Depth, achieves SotA performance on the KITTI dataset while significantly surpassing SotA on challenging, adverse condition data such as DrivingStereo, Foggy CityScape and NuScenes-Night. The project website can be found at https://kieran514.github.io/Robust-Depth-Project/.

It is essential for autonomous robots to be socially compliant while navigating in human-populated environments. Machine Learning and, especially, Deep Reinforcement Learning have recently gained considerable traction in the field of Social Navigation. This can be partially attributed to the resulting policies not being bound by human limitations in terms of code complexity or the number of variables that are handled. Unfortunately, the lack of safety guarantees and the large data requirements by DRL algorithms make learning in the real world unfeasible. To bridge this gap, simulation environments are frequently used. We propose SocNavGym, an advanced simulation environment for social navigation that can generate a wide variety of social navigation scenarios and facilitates the development of intelligent social agents. SocNavGym is lightweight, fast, easy to use, and can be effortlessly configured to generate different types of social navigation scenarios. It can also be configured to work with different hand-crafted and data-driven social reward signals and to yield a variety of evaluation metrics to benchmark agents’ performance. Further, we also provide a case study where a Dueling-DQN agent is trained to learn social-navigation policies using SocNavGym. The results provide evidence that SocNavGym can be used to train an agent from scratch to navigate in simple as well as complex social scenarios. Our experiments also show that the agents trained using the data-driven reward function display more advanced social compliance in comparison to the heuristic-based reward function.

Abstract Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to minimise the disruption caused to humans while moving. This implies predicting how people will move and complying with social conventions. Avoiding disrupting personal spaces, people’s paths and interactions are examples of these social conventions. This paper leverages Graph Neural Networks to model robot disruption considering the movement of the humans and the robot so that the model built can be used by path planning algorithms. Along with the model, this paper presents an evolution of the dataset SocNav1 (Manso et al 2020) which considers the movement of the robot and the humans, and an updated scenario-to-graph transformation which is tested using different Graph Neural Network blocks. The model trained achieves close-to-human performance in the dataset. In addition to its accuracy, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered in comparison with handcrafted models. The dataset and the model are available in a public repository ( https://github.com/gnns4hri/sngnnv2 ).

It is widely accepted that in the future, robots will cooperate with humans in everyday tasks. Robots interacting with humans will require social awareness when performing their tasks which will require navigation. While navigating, robots should aim to avoid distressing people in order to maximize their chance of social acceptance. For instance, avoiding getting too close to people or disrupting interactions. Most research approaches these problems by planning socially accepted paths, however, in everyday situations, there are many examples where a simple path planner cannot solve all of the predicted robots’ navigation problems. For instance, requesting permission to interrupt a conversation if an alternative path cannot be determined requires deliberative skills. This article presents the Social Navigation framework for Autonomous robots in Populated Environments (SNAPE), where different software agents are integrated within a robotics cognitive architecture. SNAPE addresses action planning aimed at social-awareness navigation in realistic situations: it plans socially accepted paths and conversations to negotiate its trajectory to reach targets. In this article, the framework is evaluated in different use-cases where the robot, during its navigation, has to interact with different people in order to reach its goal. The results show that participants report that the robot’s behavior was realistic and human-like.

Estimating the location and orientation of humans is an essential skill for service and assistive robots. To achieve a reliable estimation in a wide area such as an apartment, multiple RGBD cameras are frequently used. Firstly, these setups are relatively expensive. Secondly, they seldom perform an effective data fusion using the multiple camera sources at an early stage of the processing pipeline. Occlusions and partial views make this second point very relevant in these scenarios. The proposal presented in this paper makes use of graph neural networks to merge the information acquired from multiple camera sources, achieving a mean absolute error below 125mm for the location and 10° for the orientation using low-resolution RGB images. The experiments, conducted in an apartment with three cameras, benchmarked two different graph neural network implementations and a third architecture based on fully connected layers. The software used has been released as open-source in a public repository1.

Social navigation is a topic with enormous interest in autonomous robotics. Robots are gradually being used in human environments, working individually or collaborating with humans in their daily tasks. Robots in these scenarios have to be able to behave in a socially acceptable way and, for this reason, the way in which robots move has to adapt to humans and context. Proxemics has been extensively studied with the aim of improving social navigation. However, these works do not take into account that, in several situations, the personal space of the humans depends on the context (e.g., this human space is not the same in a narrow corridor than in a wide room). This work proposes the definition of an adaptive and flexible space density function that allows, on the one hand, to describe the comfort space of individuals during an interaction and, on the other hand, dynamically adapt its value in terms of the space that surrounds this interaction. In order to validate the performance, this article describes a set of simulated experiments where the robustness and improvements of the approach are tested in different environments.

Considering the widespread use of mobile robots in different parts of society, it is important to provide them with the capability to behave in a socially acceptable manner. Therefore, a research topic of great importance recently has been the study of Human-Robot Interaction. Autonomous navigation is a fundamental task in Robotics, and several different strategies that produce paths that are either length or time optimized can be found in the literature. However, considering the recent use of mobile robots in a more social context, the use of such classical techniques is restricted. Therefore, in this article we present a social navigation approach considering environments with groups of people. The proposal uses a density function to efficiently represent groups of people, and modify the navigation architecture in order to include the social behaviour of the robot during its motion. This architecture is based on the combined use of the Probabilistic Road Mapping (PRM) and the Rapidly-exploring Random Tree (RRT) path planners and an adaptation of the elastic band algorithm. Experimental evaluation was carried out in different simulated environments, providing insight on the performance of the proposed technique, which surpasses classical techniques with no proxemics awareness in terms of social impact.