Luca Maria Gambardella

Dario Mantegazza, Alessandro Giusti, Luca Maria Gambardella et al.

We consider the problem of building visual anomaly detection systems for mobile robots. Standard anomaly detection models are trained using large datasets composed only of non-anomalous data. However, in robotics applications, it is often the case that (potentially very few) examples of anomalies are available. We tackle the problem of exploiting these data to improve the performance of a Real-NVP anomaly detection model, by minimizing, jointly with the Real-NVP loss, an auxiliary outlier exposure margin loss. We perform quantitative experiments on a novel dataset (which we publish as supplementary material) designed for anomaly detection in an indoor patrolling scenario. On a disjoint test set, our approach outperforms alternatives and shows that exposing even a small number of anomalous frames yields significant performance improvements.

Dario Mantegazza, Alessandro Giusti, Luca M. Gambardella et al.

We consider the task of detecting anomalies for autonomous mobile robots based on vision. We categorize relevant types of visual anomalies and discuss how they can be detected by unsupervised deep learning methods. We propose a novel dataset built specifically for this task, on which we test a state-of-the-art approach; we finally discuss deployment in a real scenario.

Marco Ferri, Dario Mantegazza, Elia Cereda et al.

We consider the task of visually estimating the pose of a human from images acquired by a nearby nano-drone; in this context, we propose a data augmentation approach based on synthetic background substitution to learn a lightweight CNN model from a small real-world training set. Experimental results on data from two different labs proves that the approach improves generalization to unseen environments.

Umberto Junior Mele, Luca Maria Gambardella, Roberto Montemanni

Recent systems applying Machine Learning (ML) to solve the Traveling Salesman Problem (TSP) exhibit issues when they try to scale up to real case scenarios with several hundred vertices. The use of Candidate Lists (CLs) has been brought up to cope with the issues. The procedure allows to restrict the search space during solution creation, consequently reducing the solver computational burden. So far, ML were engaged to create CLs and values on the edges of these CLs expressing ML preferences at solution insertion. Although promising, these systems do not clearly restrict what the ML learns and does to create solutions, bringing with them some generalization issues. Therefore, motivated by exploratory and statistical studies, in this work we instead use a machine learning model to confirm the addition in the solution just for high probable edges. CLs of the high probable edge are employed as input, and the ML is in charge of distinguishing cases where such edges are in the optimal solution from those where they are not. . This strategy enables a better generalization and creates an efficient balance between machine learning and searching techniques. Our ML-Constructive heuristic is trained on small instances. Then, it is able to produce solutions, without losing quality, to large problems as well. We compare our results with classic constructive heuristics, showing good performances for TSPLIB instances up to 1748 cities. Although our heuristic exhibits an expensive constant time operation, we proved that the computational complexity in worst-case scenario, for the solution construction after training, is $O(n^2 \log n^2)$, being $n$ the number of vertices in the TSP instance.

Tommaso Vitali, Umberto Junior Mele, Luca Maria Gambardella et al.

The ML-Constructive heuristic is a recently presented method and the first hybrid method capable of scaling up to real scale traveling salesman problems. It combines machine learning techniques and classic optimization techniques. In this paper we present improvements to the computational weight of the original deep learning model. In addition, as simpler models reduce the execution time, the possibility of adding a local-search phase is explored to further improve performance. Experimental results corroborate the quality of the proposed improvements.

Mirko Nava, Antonio Paolillo, Jérôme Guzzi et al.

We propose a general self-supervised learning approach for spatial perception tasks, such as estimating the pose of an object relative to the robot, from onboard sensor readings. The model is learned from training episodes, by relying on: a continuous state estimate, possibly inaccurate and affected by odometry drift; and a detector, that sporadically provides supervision about the target pose. We demonstrate the general approach in three different concrete scenarios: a simulated robot arm that visually estimates the pose of an object of interest; a small differential drive robot using 7 infrared sensors to localize a nearby wall; an omnidirectional mobile robot that localizes itself in an environment from camera images. Quantitative results show that the approach works well in all three scenarios, and that explicitly accounting for uncertainty yields statistically significant performance improvements.

Daniele Palossi, Nicky Zimmerman, Alessio Burrello et al.

Artificial intelligence-powered pocket-sized air robots have the potential to revolutionize the Internet-of-Things ecosystem, acting as autonomous, unobtrusive, and ubiquitous smart sensors. With a few cm$^{2}$ form-factor, nano-sized unmanned aerial vehicles (UAVs) are the natural befit for indoor human-drone interaction missions, as the pose estimation task we address in this work. However, this scenario is challenged by the nano-UAVs' limited payload and computational power that severely relegates the onboard brain to the sub-100 mW microcontroller unit-class. Our work stands at the intersection of the novel parallel ultra-low-power (PULP) architectural paradigm and our general development methodology for deep neural network (DNN) visual pipelines, i.e., covering from perception to control. Addressing the DNN model design, from training and dataset augmentation to 8-bit quantization and deployment, we demonstrate how a PULP-based processor, aboard a nano-UAV, is sufficient for the real-time execution (up to 135 frame/s) of our novel DNN, called PULP-Frontnet. We showcase how, scaling our model's memory and computational requirement, we can significantly improve the onboard inference (top energy efficiency of 0.43 mJ/frame) with no compromise in the quality-of-result vs. a resource-unconstrained baseline (i.e., full-precision DNN). Field experiments demonstrate a closed-loop top-notch autonomous navigation capability, with a heavily resource-constrained 27-gram Crazyflie 2.1 nano-quadrotor. Compared against the control performance achieved using an ideal sensing setup, onboard relative pose inference yields excellent drone behavior in terms of median absolute errors, such as positional (onboard: 41 cm, ideal: 26 cm) and angular (onboard: 3.7$^{\circ}$, ideal: 4.1$^{\circ}$).

Dario Mantegazza, Jérôme Guzzi, Luca M. Gambardella et al.

We consider the task of controlling a quadrotor to hover in front of a freely moving user, using input data from an onboard camera. On this specific task we compare two widespread learning paradigms: a mediated approach, which learns an high-level state from the input and then uses it for deriving control signals; and an end-to-end approach, which skips high-level state estimation altogether. We show that despite their fundamental difference, both approaches yield equivalent performance on this task. We finally qualitatively analyze the behavior of a quadrotor implementing such approaches.

Mirko Nava, Jerome Guzzi, R. Omar Chavez-Garcia et al.

We introduce a general self-supervised approach to predict the future outputs of a short-range sensor (such as a proximity sensor) given the current outputs of a long-range sensor (such as a camera); we assume that the former is directly related to some piece of information to be perceived (such as the presence of an obstacle in a given position), whereas the latter is information-rich but hard to interpret directly. We instantiate and implement the approach on a small mobile robot to detect obstacles at various distances using the video stream of the robot's forward-pointing camera, by training a convolutional neural network on automatically-acquired datasets. We quantitatively evaluate the quality of the predictions on unseen scenarios, qualitatively evaluate robustness to different operating conditions, and demonstrate usage as the sole input of an obstacle-avoidance controller. We additionally instantiate the approach on a different simulated scenario with complementary characteristics, to exemplify the generality of our contribution.

Bestoun S. Ahmed, Luca M. Gambardella, Wasif Afzal et al.

Context: Combinatorial testing strategies have lately received a lot of attention as a result of their diverse applications. In its simple form, a combinatorial strategy can reduce several input parameters (configurations) of a system into a small set based on their interaction (or combination). In practice, the input configurations of software systems are subjected to constraints, especially in case of highly configurable systems. To implement this feature within a strategy, many difficulties arise for construction. While there are many combinatorial interaction testing strategies nowadays, few of them support constraints. Objective: This paper presents a new strategy, to construct combinatorial interaction test suites in the presence of constraints. Method: The design and algorithms are provided in detail. To overcome the multi-judgment criteria for an optimal solution, the multi-objective particle swarm optimization and multithreading are used. The strategy and its associated algorithms are evaluated extensively using different benchmarks and comparisons. Results: Our results are promising as the evaluation results showed the efficiency and performance of each algorithm in the strategy. The benchmarking results also showed that the strategy can generate constrained test suites efficiently as compared to state-of-the-art strategies. Conclusion: The proposed strategy can form a new way for constructing of constrained combinatorial interaction test suites. The strategy can form a new and effective base for future implementations.

Mitko Veta, Paul J. van Diest, Stefan M. Willems et al.

The proliferative activity of breast tumors, which is routinely estimated by counting of mitotic figures in hematoxylin and eosin stained histology sections, is considered to be one of the most important prognostic markers. However, mitosis counting is laborious, subjective and may suffer from low inter-observer agreement. With the wider acceptance of whole slide images in pathology labs, automatic image analysis has been proposed as a potential solution for these issues. In this paper, the results from the Assessment of Mitosis Detection Algorithms 2013 (AMIDA13) challenge are described. The challenge was based on a data set consisting of 12 training and 11 testing subjects, with more than one thousand annotated mitotic figures by multiple observers. Short descriptions and results from the evaluation of eleven methods are presented. The top performing method has an error rate that is comparable to the inter-observer agreement among pathologists.

Eduardo Feo Flushing, Luca M. Gambardella, Gianni A. Di Caro

In the context of search and rescue, we consider the problem of mission planning for heterogeneous teams that can include human, robotic, and animal agents. The problem is tackled using a mixed integer mathematical programming formulation that jointly determines the path and the activity scheduling of each agent in the team. Based on the mathematical formulation, we propose the use of soft constraints and penalties that allow the flexible strategic control of spatio-temporal relations among the search trajectories of the agents. In this way, we can enable the mission planner to obtain solutions that maximize the area coverage and, at the same time, control the spatial proximity among the agents (e.g., to minimize mutual task interference, or to promote local cooperation and data sharing). Through simulation experiments, we show the application of the strategic framework considering a number of scenarios of interest for real-world search and rescue missions.

Alessandro Giusti, Dan C. Cireşan, Jonathan Masci et al.

Deep Neural Networks now excel at image classification, detection and segmentation. When used to scan images by means of a sliding window, however, their high computational complexity can bring even the most powerful hardware to its knees. We show how dynamic programming can speedup the process by orders of magnitude, even when max-pooling layers are present.