Publications

The study of cause and effect is of the utmost importance in many branches of science, but also for many practical applications of intelligent systems. In particular, identifying causal relationships in situations that include hidden factors is a major challenge for methods that rely solely on observational data for building causal models. This article proposes CAnDOIT, a causal discovery method to reconstruct causal models using both observational and interventional time-series data. The use of interventional data in the causal analysis is crucial for real-world applications, such as robotics, where the scenario is highly complex and observational data alone are often insufficient to uncover the correct causal structure. Validation of the method is performed initially on randomly generated synthetic models and subsequently on a well-known benchmark for causal structure learning in a robotic manipulation environment. The experiments demonstrate that the approach can effectively handle data from interventions and exploit them to enhance the accuracy of the causal analysis. A Python implementation of CAnDOIT is developed and is publicly available on GitHub: https://github.com/lcastri/causalflow.

This article describes a novel technology for preserving audio documents archived on open-reel magnetic tapes forming the core of the Audio Recording Preservation (ARP) international standard. ARP is part of the Moving Picture, Audio, and Data Coding by Artificial Intelligence (MPAI) Context-based Audio Enhancement (CAE) standard, adopted by the IEEE Standard Association as IEEE 3302-2022 in December 2022. Leveraging automated Artificial Intelligence (AI) tools, ARP analyzes and extracts relevant information from digitized audio and video files of the tape’s corresponding digital Preservation Copy. This process includes identifying speed variations and surface irregularities on the tape, automatically rectifying errors to generate a restored Access Copy. By utilizing the ARP standard, archives gain a potent tool for expediting and optimizing the description of the preservation conditions of the tape, as well as automatically correcting any errors that may have occurred during the digitization process. This technology offers an efficient solution for managing both small and large collections of digitized analog items, marking a substantial advancement in the preservation of audio documents.

In SLAM (Simultaneous localization and mapping) problems, Pose Graph Optimization (PGO) is a technique to refine an initial estimate of a set of poses (positions and orientations) from a set of pairwise relative measurements. The optimization procedure can be negatively affected even by a single outlier measurement, with possible catastrophic and meaningless results. Although recent works on robust optimization aim to mitigate the presence of outlier measurements, robust solutions capable of handling large numbers of outliers are yet to come. This paper presents IPC, acronym for Incremental Probabilistic Consensus, a method that approximates the solution to the combinatorial problem of finding the maximally consistent set of measurements in an incremental fashion. It evaluates the consistency of each loop closure measurement through a consensus-based procedure, possibly applied to a subset of the global problem, where all previously integrated inlier measurements have veto power. We evaluated IPC on standard benchmarks against several state-of-the-art methods. Although it is simple and relatively easy to implement, IPC competes with or outperforms the other tested methods in handling outliers while providing online performances. We release with this paper an open-source implementation of the proposed method.

Object pose estimation is a fundamental computer vision task exploited in several robotics and augmented reality applications. Many established approaches rely on predicting 2D-3D keypoint correspondences using RANSAC (Random sample consensus) and estimating the object pose using the PnP (Perspective-n-Point) algorithm. Being RANSAC non-differentiable, correspondences cannot be directly learned in an end-to-end fashion. In this letter, we address the stereo image-based object pose estimation problem by i) introducing a differentiable RANSAC layer into a well-known monocular pose estimation network; ii) exploiting an uncertainty-driven multi-view PnP solver which can fuse information from multiple views. We evaluate our approach on a challenging public stereo object pose estimation dataset and a custom-built dataset we call Transparent Tableware Dataset (TTD), yielding state-of-the-art results against other recent approaches. Furthermore, in our ablation study, we show that the differentiable RANSAC layer plays a significant role in the accuracy of the proposed method. We release with this letter the code of our method and the TTD dataset.

Background: Re-intubation secondary to post-extubation respiratory failure in post-operative patients is associated with increased patient morbidity and mortality. Non-invasive respiratory support (NRS) alternative to conventional oxygen therapy (COT), i.e., high-flow nasal oxygen, continuous positive airway pressure, and non-invasive ventilation (NIV), has been proposed to prevent or treat post-extubation respiratory failure. Aim of the present study is assessing the effects of NRS application, compared to COT, on the re-intubation rate (primary outcome), and time to re-intubation, incidence of nosocomial pneumonia, patient discomfort, intensive care unit (ICU) and hospital length of stay, and mortality (secondary outcomes) in adult patients extubated after surgery. Methods: A systematic review and network meta-analysis of randomized and non-randomized controlled trials. A search from Medline, Embase, Scopus, Cochrane Central Register of Controlled Trials, and Web of Science from inception until February 2, 2024 was performed. Results: Thirty-three studies (11,292 patients) were included. Among all NRS modalities, only NIV reduced the re-intubation rate, compared to COT (odds ratio 0.49, 95% confidence interval 0.28; 0.87, p = 0.015, I2 = 60.5%, low certainty of evidence). In particular, this effect was observed in patients receiving NIV for treatment, while not for prevention, of post-extubation respiratory failure, and in patients at high, while not low, risk of post-extubation respiratory failure. NIV reduced the rate of nosocomial pneumonia, ICU length of stay, and ICU, hospital, and long-term mortality, while not worsening patient discomfort. Conclusions: In post-operative patients receiving NRS after extubation, NIV reduced the rate of re-intubation, compared to COT, when used for treatment of post-extubation respiratory failure and in patients at high risk of post-extubation respiratory failure.