In this paper, we focus on the problem of rendering novel views from a Neural Radiance Field (NeRF) under unobserved light conditions. To this end, we introduce a novel dataset, dubbed ReNe (Relighting NeRF), framing real world objects under one-light-at-time (OLAT) conditions, annotated with accurate ground-truth camera and light poses. Our acquisition pipeline leverages two robotic arms holding, respectively, a camera and an omni-directional point-wise light source. We release a total of 20 scenes depicting a variety of objects with complex geometry and challenging materials. Each scene includes 2000 images, acquired from 50 different points of views under 40 different OLAT conditions. By leveraging the dataset, we perform an ablation study on the relighting capability of variants of the vanilla NeRF architecture and identify a lightweight architecture that can render novel views of an object under novel light conditions, which we use to establish a non-trivial baseline for the dataset. Dataset and benchmark are available at https://eyecan-ai.

We introduce a novel framework for training deep stereo networks effortlessly and without any ground-truth. By leveraging state-of-the-art neural rendering solutions, we generate stereo training data from image sequences collected with a single handheld camera. On top of them, a NeRF-supervised training procedure is carried out, from which we exploit rendered stereo triplets to compensate for occlusions and depth maps as proxy labels. This results in stereo networks capable of predicting sharp and detailed disparity maps. Experimental results show that models trained under this regime yield a 30-40% improvement over existing self-supervised methods on the challenging Middle-bury dataset, filling the gap to supervised models and, most times, outperforming them at zero-shot generalization.

In this paper, we propose the first-ever real benchmark thought for evaluating Neural Radiance Fields (NeRFs) and, in general, Neural Rendering (NR) frameworks. We design and implement an effective pipeline for scanning real objects in quantity and effortlessly. Our scan station is built with less than 500$ hardware budget and can collect roughly 4000 images of a scanned object in just 5 minutes. Such a platform is used to build ScanNeRF, a dataset characterized by several train/val/test splits aimed at benchmarking the performance of modern NeRF methods under different conditions. Accordingly, we evaluate three cuttingedge NeRF variants on it to highlight their strengths and weaknesses. The dataset is available on our project page, together with an online benchmark to foster the development of better and better NeRFs.

In this article, an innovative algorithm for instance segmentation of wires called Ariadne+ is presented. Although vastly present in many manufacturing environments, the perception and manipulation of wires is still an open problem for robotic applications. Wires are deformable linear objects lacking of any specific shape, color, and feature. The proposed approach uses deep learning and standard computer vision techniques aiming at their reliable and time effective instance segmentation. A deep convolutional neural network is employed to generate a binary mask showing where wires are present in the input image, then the graph theory is applied to create the wire paths from the binary mask through an iterative approach that aims to maximize the graph coverage. In addition, the B-Spline model of each instance, useful in manipulation tasks, is provided. The approach has been validated quantitatively and qualitatively using a manually labeled test dataset and by comparing it against the original Ariadne algorithm. The timings performances of the approach have been also analyzed in depth.

In this paper we investigate how to effectively deploy deep learning in practical industrial settings, such as robotic grasping applications. When a deep-learning based solution is proposed, usually lacks of any simple method to generate the training data. In the industrial field, where automation is the main goal, not bridging this gap is one of the main reasons why deep learning is not as widespread as it is in the academic world. For this reason, in this work we developed a system composed by a 3-DoF Pose Estimator based on Convolutional Neural Networks (CNNs) and an effective procedure to gather massive amounts of training images in the field with minimal human intervention. By automating the labeling stage, we also obtain very robust systems suitable for production-level usage. An open source implementation of our solution is provided, alongside with the dataset used for the experimental evaluation.

In this article, we propose an augmented reality semiautomatic labeling (ARS), a semiautomatic method which leverages on moving a 2-D camera by means of a robot, proving precise camera tracking, and an augmented reality pen (ARP) to define initial object bounding box, to create large labeled data sets with minimal human intervention. By removing the burden of generating annotated data from humans, we make the deep learning technique applied to computer vision, which typically requires very large data sets, truly automated and reliable. With the ARS pipeline, we created two novel data sets effortlessly, one on electromechanical components (industrial scenario) and other on fruits (daily-living scenario) and trained two state-of-the-art object detectors robustly, based on convolutional neural networks, such as you only look once (YOLO) and single shot detector (SSD). With respect to conventional manual annotation of 1000 frames that takes us slightly more than 10 h, the proposed approach based on ARS allows to annotate 9 sequences of about 35 000 frames in less than 1 h, with a gain factor of about 450. Moreover, both the precision and recall of object detection is increased by about 15% with respect to manual labeling. All our software is available as a robot operating system (ROS) package in a public repository alongside with the novel annotated data sets. Note to Practitioners-This article was motivated by the lack of a simple and effective solution for the generation of data sets usable to train a data-driven model, such as a modern deep neural network, so as to make them accessible in an industrial environment. Specifically, a deep learning robot guidance vision system would require such a large amount of manually labeled images that it would be too expensive and impractical for a real use case, where system reconfigurability is a fundamental requirement. With our system, on the other hand, especially in the field of industrial robotics, the cost of image labeling can be reduced, for the first time, to nearly zero, thus paving the way for self-reconfiguring systems with very high performance (as demonstrated by our experimental results). One of the limitations of this approach is the need to use a manual method for the detection of objects of interest in the preliminary stages of the pipeline (ARP or graphical interface). A feasible extension, related to the field of collaborative robotics, could be used to exploit the robot itself, manually moved by the user, even for this preliminary stage, so as to eliminate any source of inaccuracy.

In this paper, a tactile-based robotic system to perform the insertion of a Deformable Linear Object (DLO) in a hole is proposed. This is a typical application in manufacturing processes involving assembly of electric cables in connectors or electromechanical components. A Recurrent Neural Network (RNN) with Long Short Term Memory (LSTM) cells is adopted in this work to predict the external forces acting on the DLO from the tactile data. The tactile sensor, mounted on the finger, provides 16 signals and it is the only sensor required during the effective insertion task. In a real environment, the tight spaces very often prevent the possibility to use the vision system, also when the same task is performed by a human being. Force/Torque sensors instead increase the system price and provide signals that might be affected by inertia disturbance or other undesired effect, that are difficult to manage. The control law design is based on the RNN outputs and the distance between end-effector and target hole. In particular, it leads the plastically deformed DLO inside the hole while it adjusts the tool pose guided by the control errors in order to prevent buckling. The contribute of this work is dual: first a tactile feedback integrating a RNN to estimate contact forces on the grasped object is presented; second, a control system is developed to perform a challenging insertion of a DLO in a hole. Experimental works are presented to validate the proposed algorithms.

This paper reports the development of a manipulation system for electric wires, implemented by means of a commercial gripper installed on an industrial manipulator and equipped with cameras and suitably designed tactile sensors. The purpose of this system is the execution of wire insertion on commercial electromechanical components. The synergy between computer vision and tactile sensing is necessary because, in a real environment, the tight spaces very often prevent the possibility to use the vision system, also when the same task is performed by a human being. A novel technique to speed up the generation of training data sets for convolutional neural networks (CNNs) is proposed. Therefore, this technique is used to train a CNN in order to detect small objects (such as wire terminals). Moreover, aiming to prevent faults during the task and to interact with the environment safely, several machine learning approaches are used to produce an affordable output from the tactile sensor. The proposed approach shows how a cheap sensor embedded with suitable intelligence can provide information comparable to a more expensive force sensor.

We present a novel mapping framework for robot navigation which features a multi-level querying system capable to obtain rapidly representations as diverse as a 3D voxel grid, a 2.5D height map and a 2D occupancy grid. These are inherently embedded into a memory and time efficient core data structure organized as a Tree of SkipLists. Compared to the well-known Octree representation, our approach exhibits a better time efficiency, thanks to its simple and highly parallelizable computational structure, and a similar memory footprint when mapping large workspaces. Peculiarly within the realm of mapping for robot navigation, our framework supports real-time erosion and re-integration of measurements upon reception of optimized poses from the sensor tracker, so as to improve continuously the accuracy of the map.