Luis J. Manso

Anthony Francis, Claudia Pérez-D'Arpino, Chengshu Li et al. · cmu, mit

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 paper, we pave the road towards 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.

Kieran Saunders, George Vogiatzis, Luis J. Manso

Self-supervised monocular depth estimation has been a subject of intense study in recent years, because of its applications in robotics and autonomous driving. Much of the recent work focuses on improving depth estimation by increasing architecture complexity. This paper shows that state-of-the-art performance can also be achieved by improving the learning process rather than increasing model complexity. More specifically, we propose (i) disregarding small potentially dynamic objects when training, and (ii) employing an appearance-based approach to separately estimate object pose for truly dynamic objects. We demonstrate that these simplifications reduce GPU memory usage by 29% and result in qualitatively and quantitatively improved depth maps. The code is available at https://github.com/kieran514/Dyna-DM.

Kieran Saunders, George Vogiatzis, Luis Manso

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 here https://kieran514.github.io/Robust-Depth-Project/.

Daniel Rodriguez-Criado, Pilar Bachiller, George Vogiatzis et al.

Its numerous applications make multi-human 3D pose estimation a remarkably impactful area of research. Nevertheless, assuming a multiple-view system composed of several regular RGB cameras, 3D multi-pose estimation presents several challenges. First of all, each person must be uniquely identified in the different views to separate the 2D information provided by the cameras. Secondly, the 3D pose estimation process from the multi-view 2D information of each person must be robust against noise and potential occlusions in the scenario. In this work, we address these two challenges with the help of deep learning. Specifically, we present a model based on Graph Neural Networks capable of predicting the cross-view correspondence of the people in the scenario along with a Multilayer Perceptron that takes the 2D points to yield the 3D poses of each person. These two models are trained in a self-supervised manner, thus avoiding the need for large datasets with 3D annotations.

Aditya Kapoor, Sushant Swamy, Luis Manso et al.

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 light-weight, 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 provides 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 displays more advanced social compliance in comparison to the heuristic-based reward function.

Vincent Gbouna Zakka, Zhuangzhuang Dai, Luis J. Manso

The growing ageing population and their preference to maintain independence by living in their own homes require proactive strategies to ensure safety and support. Ambient Assisted Living (AAL) technologies have emerged to facilitate ageing in place by offering continuous monitoring and assistance within the home. Within AAL technologies, action recognition plays a crucial role in interpreting human activities and detecting incidents like falls, mobility decline, or unusual behaviours that may signal worsening health conditions. However, action recognition in practical AAL applications presents challenges, including occlusions, noisy data, and the need for real-time performance. While advancements have been made in accuracy, robustness to noise, and computation efficiency, achieving a balance among them all remains a challenge. To address this challenge, this paper introduces the Robust and Efficient Temporal Convolution network (RE-TCN), which comprises three main elements: Adaptive Temporal Weighting (ATW), Depthwise Separable Convolutions (DSC), and data augmentation techniques. These elements aim to enhance the model's accuracy, robustness against noise and occlusion, and computational efficiency within real-world AAL contexts. RE-TCN outperforms existing models in terms of accuracy, noise and occlusion robustness, and has been validated on four benchmark datasets: NTU RGB+D 60, Northwestern-UCLA, SHREC'17, and DHG-14/28. The code is publicly available at: https://github.com/Gbouna/RE-TCN

Kieran Saunders, Luis J. Manso, George Vogiatzis

In the domain of multi-baseline stereo, the conventional understanding is that, in general, increasing baseline separation substantially enhances the accuracy of depth estimation. However, prevailing self-supervised depth estimation architectures primarily use minimal frame separation and a constrained stereo baseline. Larger frame separations can be employed; however, we show this to result in diminished depth quality due to various factors, including significant changes in brightness, and increased areas of occlusion. In response to these challenges, our proposed method, BaseBoostDepth, incorporates a curriculum learning-inspired optimization strategy to effectively leverage larger frame separations. However, we show that our curriculum learning-inspired strategy alone does not suffice, as larger baselines still cause pose estimation drifts. Therefore, we introduce incremental pose estimation to enhance the accuracy of pose estimations, resulting in significant improvements across all depth metrics. Additionally, to improve the robustness of the model, we introduce error-induced reconstructions, which optimize reconstructions with added error to the pose estimations. Ultimately, our final depth network achieves state-of-the-art performance on KITTI and SYNS-patches datasets across image-based, edge-based, and point cloud-based metrics without increasing computational complexity at test time. The project website can be found at https://kieran514.github.io/BaseBoostDepth-Project.

Zhuangzhuang Dai, Zhongxi Lu, Vincent G. Zakka et al.

Estimating human gaze target from visible images is a critical task for robots to understand human attention, yet the development of generalizable neural architectures and training paradigms remains challenging. While recent advances in pre-trained vision foundation models offer promising avenues for locating gaze targets, the integration of multi-modal cues -- including eyes, head poses, gestures, and contextual features -- demands adaptive and efficient decoding mechanisms. Inspired by Mixture-of-Experts (MoE) for adaptive domain expertise in large vision-language models, we propose GazeMoE, a novel end-to-end framework that selectively leverages gaze-target-related cues from a frozen foundation model through MoE modules. To address class imbalance in gaze target classification (in-frame vs. out-of-frame) and enhance robustness, GazeMoE incorporates a class-balancing auxiliary loss alongside strategic data augmentations, including region-specific cropping and photometric transformations. Extensive experiments on benchmark datasets demonstrate that our GazeMoE achieves state-of-the-art performance, outperforming existing methods on challenging gaze estimation tasks. The code and pre-trained models are released at https://huggingface.co/zdai257/GazeMoE

Zhuangzhuang Dai, Vincent Gbouna Zakka, Luis J. Manso et al.

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.

Vincent Gbouna Zakka, Luis J. Manso, Zhuangzhuang Dai

Human activity recognition is increasingly vital for supporting independent living, particularly for the elderly and those in need of assistance. Domestic service robots with monitoring capabilities can enhance safety and provide essential support. Although image-based methods have advanced considerably in the past decade, their adoption remains limited by concerns over privacy and sensitivity to low-light or dark conditions. As an alternative, millimetre-wave (mmWave) radar can produce point cloud data which is privacy-preserving. However, processing the sparse and noisy point clouds remains a long-standing challenge. While graph-based methods and attention mechanisms show promise, they predominantly rely on "fixed" kernels; kernels that are applied uniformly across all neighbourhoods, highlighting the need for adaptive approaches that can dynamically adjust their kernels to the specific geometry of each local neighbourhood in point cloud data. To overcome this limitation, we introduce an adaptive approach within the graph convolutional framework. Instead of a single shared weight function, our Multi-Head Adaptive Kernel (MAK) module generates multiple dynamic kernels, each capturing different aspects of the local feature space. By progressively refining local features while maintaining global spatial context, our method enables convolution kernels to adapt to varying local features. Experimental results on benchmark datasets confirm the effectiveness of our approach, achieving state-of-the-art performance in human activity recognition. Our source code is made publicly available at: https://github.com/Gbouna/MAK-GCN

Pilar Bachiller, Daniel Rodriguez-Criado, Ronit R. Jorvekar et al.

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 [25] 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).

Daniel Rodriguez-Criado, Pilar Bachiller, Pablo Bustos et al.

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 125 mm for the location and 10 degrees 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 repository (https://github.com/vangiel/WheresTheFellow).

Daniel Rodriguez-Criado, Maria Chli, Luis J. Manso et al.

Traffic congestion in urban areas presents significant challenges, and Intelligent Transportation Systems (ITS) have sought to address these via automated and adaptive controls. However, these systems often struggle to transfer simulated experiences to real-world scenarios. This paper introduces a novel methodology for bridging this `sim-real' gap by creating photorealistic images from 2D traffic simulations and recorded junction footage. We propose a novel image generation approach, integrating a Conditional Generative Adversarial Network with a Graph Neural Network (GNN) to facilitate the creation of realistic urban traffic images. We harness GNNs' ability to process information at different levels of abstraction alongside segmented images for preserving locality data. The presented architecture leverages the power of SPADE and Graph ATtention (GAT) network models to create images based on simulated traffic scenarios. These images are conditioned by factors such as entity positions, colors, and time of day. The uniqueness of our approach lies in its ability to effectively translate structured and human-readable conditions, encoded as graphs, into realistic images. This advancement contributes to applications requiring rich traffic image datasets, from data augmentation to urban traffic solutions. We further provide an application to test the model's capabilities, including generating images with manually defined positions for various entities.

Jordan J. Bird, Chloe M. Barnes, Luis J. Manso et al.

Contemporary Artificial Intelligence technologies allow for the employment of Computer Vision to discern good crops from bad, providing a step in the pipeline of selecting healthy fruit from undesirable fruit, such as those which are mouldy or gangrenous. State-of-the-art works in the field report high accuracy results on small datasets (<1000 images), which are not representative of the population regarding real-world usage. The goals of this study are to further enable real-world usage by improving generalisation with data augmentation as well as to reduce overfitting and energy usage through model pruning. In this work, we suggest a machine learning pipeline that combines the ideas of fine-tuning, transfer learning, and generative model-based training data augmentation towards improving fruit quality image classification. A linear network topology search is performed to tune a VGG16 lemon quality classification model using a publicly-available dataset of 2690 images. We find that appending a 4096 neuron fully connected layer to the convolutional layers leads to an image classification accuracy of 83.77%. We then train a Conditional Generative Adversarial Network on the training data for 2000 epochs, and it learns to generate relatively realistic images. Grad-CAM analysis of the model trained on real photographs shows that the synthetic images can exhibit classifiable characteristics such as shape, mould, and gangrene. A higher image classification accuracy of 88.75% is then attained by augmenting the training with synthetic images, arguing that Conditional Generative Adversarial Networks have the ability to produce new data to alleviate issues of data scarcity. Finally, model pruning is performed via polynomial decay, where we find that the Conditional GAN-augmented classification network can retain 81.16% classification accuracy when compressed to 50% of its original size.

Daniel Rodriguez-Criado, Pilar Bachiller, Luis J. Manso

Minimising the discomfort caused by robots when navigating in social situations is crucial for them to be accepted. The paper presents a machine learning-based framework that bootstraps existing one-dimensional datasets to generate a cost map dataset and a model combining Graph Neural Network and Convolutional Neural Network layers to produce cost maps for human-aware navigation in real-time. The proposed framework is evaluated against the original one-dimensional dataset and in simulated navigation tasks. The results outperform similar state-of-the-art-methods considering the accuracy on the dataset and the navigation metrics used. The applications of the proposed framework are not limited to human-aware navigation, it could be applied to other fields where map generation is needed.

Rishabh Baghel, Aditya Kapoor, Pilar Bachiller et al.

Social navigation datasets are necessary to assess social navigation algorithms and train machine learning algorithms. Most of the currently available datasets target pedestrians' movements as a pattern to be replicated by robots. It can be argued that one of the main reasons for this to happen is that compiling datasets where real robots are manually controlled, as they would be expected to behave when moving, is a very resource-intensive task. Another aspect that is often missing in datasets is symbolic information that could be relevant, such as human activities, relationships or interactions. Unfortunately, the available datasets targeting robots and supporting symbolic information are restricted to static scenes. This paper argues that simulation can be used to gather social navigation data in an effective and cost-efficient way and presents a toolkit for this purpose. A use case studying the application of graph neural networks to create learned control policies using supervised learning is presented as an example of how it can be used.

Luis J. Manso, Ronit R. Jorvekar, Diego R. Faria et al.

Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to navigate avoiding going through the personal spaces of the people surrounding them. Complying with social rules such as not getting in the middle of human-to-human and human-to-object interactions is also important. This paper suggests using Graph Neural Networks to model how inconvenient the presence of a robot would be in a particular scenario according to learned human conventions so that it can be used by path planning algorithms. To do so, we propose two ways of modelling social interactions using graphs and benchmark them with different Graph Neural Networks using the SocNav1 dataset. We achieve close-to-human performance in the dataset and argue that, in addition to promising results, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered and easily embedded in code, in comparison with model-based approaches. The code used to train and test the resulting graph neural network is available in a public repository.

A. Romero-Garces, L. J. Manso, Marco A. Gutierez et al.

There is currently a large amount of robotics software using the component-oriented programming paradigm. However, the rapid growth in number and complexity of components may compromise the scalability and the whole lifecycle of robotics software systems. Model-Driven Engineering can be used to mitigate these problems. This paper describes how using Domain-Specific Languages to generate and describe critical parts of robotic systems helps developers to perform component managerial tasks such as component creation, modification, monitoring and deployment. Four different DSLs are proposed in this paper: i) CDSL for specifying the structure of the components, ii) IDSL for the description of their interfaces, iii) DDSL for describing the deployment process of component networks and iv) PDSL to define and configure component parameters. Their benefits have been demonstrated after their implementation in RoboComp, a general-purpose and component-based robotics framework. Examples of the usage of these DSLs are shown along with experiments that demonstrate the benefits they bring to the lifecycle of the components.