Giulia De Masi - Academia.edu (original) (raw)

Papers by Giulia De Masi

arXiv (Cornell University), Apr 8, 2022

The goal of this work is to apply a denoising image transformer to remove the distortion from und... more The goal of this work is to apply a denoising image transformer to remove the distortion from underwater images and compare it with other similar approaches. Automatic restoration of underwater images plays an important role since it allows to increase the quality of the images, without the need for more expensive equipment. This is a critical example of the important role of the machine learning algorithms to support marine exploration and monitoring, reducing the need for human intervention like the manual processing of the images, thus saving time, effort, and cost. This paper is the first application of the image transformer-based approach called "Pre-Trained Image Processing Transformer" to underwater images. This approach is tested on the UFO-120 dataset, containing 1500 images with the corresponding clean images.

Journal of Economic Interaction and Coordination, Jul 16, 2023

Globalization has considerably increased the movement of people and goods around the world, which... more Globalization has considerably increased the movement of people and goods around the world, which constitutes a key channel of viral infection. Increasingly close economic links between countries speeds up the transfer of goods and information, and the knock-on effect of economic crises, but also the transmission of diseases. Foreign direct investment (FDI) and trade establish clear ties between countries of origin and destination, and it is along these chains that contagious phenomena can unfold. In this paper, we investigate whether countries' centrality in both global production and trade network corresponds to higher COVID-19 infection and mortality rates. Merging data on EU-27 greenfield FDI and international trade with data on COVID-19 infections and deaths, we find that countries mostly exposed to the COVID-19 outbreak are those characterized by a higher eigenvector centrality. This result is robust to the use of an alternative measure of network centrality and to the inclusion of other possible confounding factors.

The Twentieth International Offshore and Polar Engineering Conference, Jun 20, 2010

In this paper the Artificial Neural Networks (ANN) are used to produce a reliable short term (up ... more In this paper the Artificial Neural Networks (ANN) are used to produce a reliable short term (up to 100 s) prediction of the actual sea surface elevation, named "wave nowcasting". This is of primary importance to support marine operations of vessels in Offshore Engineering. The present study proposes a combined application of ANN and wavegrouping analysis based on the Hilbert transform. The observations of empirical time series are used to verify the performance of this model and to find the best configuration. The results indicate that the radial basis function networks provide an effective and accurate tool to forecast the arrival of wave groups at the vessel location.

Proceedings of SPIE, Jun 7, 2007

We use the theory of complex networks in order to quantitatively characterize the structure of re... more We use the theory of complex networks in order to quantitatively characterize the structure of reciprocal expositions of Italian banks in the interbank money market market. We observe two main different strategies of banks: small banks tend to be the lender of the system, while large banks are borrowers. We propose a model to reproduce the main statistical features of this market. Moreover the network analysis allows us to investigate properties of robustness of this system. * e-MID is run by e-MID S.p.A. Società Interbancaria per l'Automazione (SIA), Milan. The central system is located in the office of the SIA and the peripherals on the premises of the member participants.

arXiv (Cornell University), Jan 15, 2009

An analysis of the Japanese credit market in 2004 between banks and quoted firms is done in this ... more An analysis of the Japanese credit market in 2004 between banks and quoted firms is done in this paper using the tools of the networks theory. It can be pointed out that: (i) a backbone of the credit channel emerges, where some links play a crucial role; (ii) big banks privilege long-term contracts; the "minimal spanning trees" (iii) disclose a highly hierarchical backbone, where the central positions are occupied by the largest banks, and emphasize (iv) a strong geographical characterization, while (v) the clusters of firms do not have specific common properties. Moreover, (vi) while larger firms have multiple lending in large, (vii) the demand for credit (long vs. short term debt and multi-credit lines) of firms with similar sizes is very heterogeneous.

2023 IEEE International Conference on Robotics and Automation (ICRA)

arXiv (Cornell University), Feb 2, 2023

Spiking neural networks are becoming increasingly popular for their low energy requirement in rea... more Spiking neural networks are becoming increasingly popular for their low energy requirement in real-world tasks with accuracy comparable to the traditional ANNs. SNN training algorithms face the loss of gradient information and nondifferentiability due to the Heaviside function in minimizing the model loss over model parameters. To circumvent the problem surrogate method uses a differentiable approximation of the Heaviside in the backward pass, while the forward pass uses the Heaviside as the spiking function. We propose to use the zeroth order technique at the neuron level to resolve this dichotomy and use it within the automatic differentiation tool. As a result, we establish a theoretical connection between the proposed local zeroth-order technique and the existing surrogate methods and vice-versa. The proposed method naturally lends itself to energyefficient training of SNNs on GPUs. Experimental results with neuromorphic datasets show that such implementation requires less than 1% neurons to be active in the backward pass, resulting in a 100x speed-up in the backward computation time. Our method offers better generalization compared to the state-of-the-art energy-efficient technique while maintaining similar efficiency.

Springer eBooks, 2022

The goal of this work is to apply a denoising image transformer to remove the distortion from und... more The goal of this work is to apply a denoising image transformer to remove the distortion from underwater images and compare it with other similar approaches. Automatic restoration of underwater images plays an important role since it allows to increase the quality of the images, without the need for more expensive equipment. This is a critical example of the important role of the machine learning algorithms to support marine exploration and monitoring, reducing the need for human intervention like the manual processing of the images, thus saving time, effort, and cost. This paper is the first application of the image transformer-based approach called "Pre-Trained Image Processing Transformer" to underwater images. This approach is tested on the UFO-120 dataset, containing 1500 images with the corresponding clean images.

Ocean Engineering, 2019

In Engineering design, it is often necessary to calculate cyclonic extremes such as winds and wav... more In Engineering design, it is often necessary to calculate cyclonic extremes such as winds and waves associated with certain exceedance probabilities. Estimation is difficult, however, for at least two reasons. The first is related to recorded data mainly referring to previous decades. The estimate of extremes over very long periods is heavily influenced by the small number of severe events that have occurred in recent history. The second concerns the spatial variability of recorded cyclone tracks. Because of the low frequency of cyclone occurrence, estimates of extremes made from a limited database can vary substantially over relatively small distances, even within a spatially limited region where it would be reasonable to expect homogeneous values. Statistical uncertainty grows exponentially, moving toward cyclonic peripheral areas affected very unfrequently by cyclone passages. The commonly used methods to account for this uncertainty are historical track shifting and deductive approaches, but a certain degree of subjectivity is intrinsic in these approaches. For this reason, the aim of this paper is to build a more proper statistical methodology to account for cyclone spatial variability on extreme evaluations. It is based on the Inverse First Order Reliability Method (IFORM) used to determine the bivariate exceedance probability of the wind speed/track distance for cyclonic events in any selected location. By means of numerical modelling, namely Simulating WAves Nearshore (SWAN), the proposed approach, through a logical sequence, predicts the cyclonic track characteristics for a certain return period which, at the selected location, induce the worst conditions in terms of maximum sea state during a cyclonic event. Compared with the historical methods, which consider the effects of all the cyclones occurring in the area, the method is able to account for spatial variability even within regions with a very low cyclonic risk. The methodology produces consistent results, avoiding spurious discontinuities between contiguous locations which presumably have very similar vulnerability to cyclones.

Journal of Physics: Conference Series, 2017

Corrosion is one of the principal causes of degradation to failure of marine structures. In pract... more Corrosion is one of the principal causes of degradation to failure of marine structures. In practice, localized corrosion is the most dangerous mode of attack and can result in serious failures, in particular in marine flowlines and inter-field lines, arousing serious concerns relatively to environmental impact. The progress in time of internal corrosion, the location along the route and across the pipe section, the development pattern and the depth of the loss of metal are a very complex issue: the most important factors are products characteristics, transport conditions over the operating lifespan, process fluid-dynamics, and pipeline geometrical configuration. Understanding which factors among them play the most important role is a key step to develop a model able to predict with enough accuracy the sections more exposed to risk of failure. Some factors play a crucial role at certain spatial scales while other factors at other scales. The Mutual Information Theory, intimately related to the concept of Shannon Entropy in Information theory, has been applied to detect the most important variables at each scale. Finally, the variables emerged from this analysis at each scale have been integrated in a predicting data driven model sensibly improving its performance.

arXiv (Cornell University), Nov 17, 2022

Strongly opinionated minorities can have a dramatic impact on the opinion dynamics of a large pop... more Strongly opinionated minorities can have a dramatic impact on the opinion dynamics of a large population. Two factions of inflexible minorities, polarised into two competing opinions, could lead the entire population to persistent indecision. Equivalently, populations can remain undecided when individuals sporadically change their opinion based on individual information rather than social information. Our analysis compares the cross-inhibition model with the voter model for decisions between equally good alternatives, and with the weighted voter model for decisions among alternatives characterised by different qualities. Here we show that cross-inhibition, contrary to the other two models, is a simple mechanism that allows the population to reach a stable majority for one alternative even in the presence of a relatively high amount of asocial behaviour. The results predicted by the mean-field models are confirmed by experiments with swarms of 100 locally interacting robots. This work suggests an answer to the longstanding question of why inhibitory signals are widespread in natural systems of collective decision making, and, at the same time, it proposes an efficient mechanism for designing resilient swarms of minimalistic robots.

IEEE Transactions on Wireless Communications, 2023

In this paper, a Universal Underwater Software Defined Modem (UniSDM) architecture is proposed th... more In this paper, a Universal Underwater Software Defined Modem (UniSDM) architecture is proposed that may operate in different modes (acoustic, magnetic induction, optical and RF), in order to utilize the advantages of each mode and accordingly satisfy the requirements of many latest use cases in underwater communication systems. A detailed description of the novel UniSDM architecture is presented first. The novelty of this architecture is its flexibility, i.e., allowing the designers to produce a device that may include any type of modes operating seamlessly and jointly by exchanging data, control and synchronization. Many challenges, including high system costs and coordination between different modes, are addressed in the paper. Moreover, numerical evaluation is conducted to assess the performance of the proposed UniSDM architecture. Finally, the performance evaluation shows that the utilization of the UniSDM allows to decrease the transmission latency and improve the energy efficiency, while maintaining high reliability and robustness in underwater communication systems.

OCEANS 2022, Hampton Roads

OCEANS 2022 - Chennai

Underwater manipulation is one of the most remarkable ongoing research subjects in robotics. Inte... more Underwater manipulation is one of the most remarkable ongoing research subjects in robotics. Intervention autonomous underwater vehicles (I-AUVs) not only have to cope with the technical challenges associated with traditional manipulation tasks but do so while currents and waves perturb the stability of the vehicle, and low-light, turbid water conditions impede perceiving the surroundings. Certainly, the dynamic nature and our limited understanding of the marine environment hinder the autonomous performance of underwater robot manipulation. This manuscript provides a discussion on previous research and the limiting factors that impose on the longenvisioned prospects of autonomous underwater manipulation to finally highlight research directions that have the potential to improve the autonomy capabilities of I-AUVs.

GLOBECOM 2022 - 2022 IEEE Global Communications Conference, Dec 4, 2022

arXiv (Cornell University), Apr 8, 2022

The goal of this work is to apply a denoising image transformer to remove the distortion from und... more The goal of this work is to apply a denoising image transformer to remove the distortion from underwater images and compare it with other similar approaches. Automatic restoration of underwater images plays an important role since it allows to increase the quality of the images, without the need for more expensive equipment. This is a critical example of the important role of the machine learning algorithms to support marine exploration and monitoring, reducing the need for human intervention like the manual processing of the images, thus saving time, effort, and cost. This paper is the first application of the image transformer-based approach called "Pre-Trained Image Processing Transformer" to underwater images. This approach is tested on the UFO-120 dataset, containing 1500 images with the corresponding clean images.

Journal of Economic Interaction and Coordination, Jul 16, 2023

Globalization has considerably increased the movement of people and goods around the world, which... more Globalization has considerably increased the movement of people and goods around the world, which constitutes a key channel of viral infection. Increasingly close economic links between countries speeds up the transfer of goods and information, and the knock-on effect of economic crises, but also the transmission of diseases. Foreign direct investment (FDI) and trade establish clear ties between countries of origin and destination, and it is along these chains that contagious phenomena can unfold. In this paper, we investigate whether countries' centrality in both global production and trade network corresponds to higher COVID-19 infection and mortality rates. Merging data on EU-27 greenfield FDI and international trade with data on COVID-19 infections and deaths, we find that countries mostly exposed to the COVID-19 outbreak are those characterized by a higher eigenvector centrality. This result is robust to the use of an alternative measure of network centrality and to the inclusion of other possible confounding factors.

The Twentieth International Offshore and Polar Engineering Conference, Jun 20, 2010

In this paper the Artificial Neural Networks (ANN) are used to produce a reliable short term (up ... more In this paper the Artificial Neural Networks (ANN) are used to produce a reliable short term (up to 100 s) prediction of the actual sea surface elevation, named "wave nowcasting". This is of primary importance to support marine operations of vessels in Offshore Engineering. The present study proposes a combined application of ANN and wavegrouping analysis based on the Hilbert transform. The observations of empirical time series are used to verify the performance of this model and to find the best configuration. The results indicate that the radial basis function networks provide an effective and accurate tool to forecast the arrival of wave groups at the vessel location.

Proceedings of SPIE, Jun 7, 2007

We use the theory of complex networks in order to quantitatively characterize the structure of re... more We use the theory of complex networks in order to quantitatively characterize the structure of reciprocal expositions of Italian banks in the interbank money market market. We observe two main different strategies of banks: small banks tend to be the lender of the system, while large banks are borrowers. We propose a model to reproduce the main statistical features of this market. Moreover the network analysis allows us to investigate properties of robustness of this system. * e-MID is run by e-MID S.p.A. Società Interbancaria per l'Automazione (SIA), Milan. The central system is located in the office of the SIA and the peripherals on the premises of the member participants.

arXiv (Cornell University), Jan 15, 2009

An analysis of the Japanese credit market in 2004 between banks and quoted firms is done in this ... more An analysis of the Japanese credit market in 2004 between banks and quoted firms is done in this paper using the tools of the networks theory. It can be pointed out that: (i) a backbone of the credit channel emerges, where some links play a crucial role; (ii) big banks privilege long-term contracts; the "minimal spanning trees" (iii) disclose a highly hierarchical backbone, where the central positions are occupied by the largest banks, and emphasize (iv) a strong geographical characterization, while (v) the clusters of firms do not have specific common properties. Moreover, (vi) while larger firms have multiple lending in large, (vii) the demand for credit (long vs. short term debt and multi-credit lines) of firms with similar sizes is very heterogeneous.

2023 IEEE International Conference on Robotics and Automation (ICRA)

arXiv (Cornell University), Feb 2, 2023

Spiking neural networks are becoming increasingly popular for their low energy requirement in rea... more Spiking neural networks are becoming increasingly popular for their low energy requirement in real-world tasks with accuracy comparable to the traditional ANNs. SNN training algorithms face the loss of gradient information and nondifferentiability due to the Heaviside function in minimizing the model loss over model parameters. To circumvent the problem surrogate method uses a differentiable approximation of the Heaviside in the backward pass, while the forward pass uses the Heaviside as the spiking function. We propose to use the zeroth order technique at the neuron level to resolve this dichotomy and use it within the automatic differentiation tool. As a result, we establish a theoretical connection between the proposed local zeroth-order technique and the existing surrogate methods and vice-versa. The proposed method naturally lends itself to energyefficient training of SNNs on GPUs. Experimental results with neuromorphic datasets show that such implementation requires less than 1% neurons to be active in the backward pass, resulting in a 100x speed-up in the backward computation time. Our method offers better generalization compared to the state-of-the-art energy-efficient technique while maintaining similar efficiency.

Springer eBooks, 2022

The goal of this work is to apply a denoising image transformer to remove the distortion from und... more The goal of this work is to apply a denoising image transformer to remove the distortion from underwater images and compare it with other similar approaches. Automatic restoration of underwater images plays an important role since it allows to increase the quality of the images, without the need for more expensive equipment. This is a critical example of the important role of the machine learning algorithms to support marine exploration and monitoring, reducing the need for human intervention like the manual processing of the images, thus saving time, effort, and cost. This paper is the first application of the image transformer-based approach called "Pre-Trained Image Processing Transformer" to underwater images. This approach is tested on the UFO-120 dataset, containing 1500 images with the corresponding clean images.

Ocean Engineering, 2019

In Engineering design, it is often necessary to calculate cyclonic extremes such as winds and wav... more In Engineering design, it is often necessary to calculate cyclonic extremes such as winds and waves associated with certain exceedance probabilities. Estimation is difficult, however, for at least two reasons. The first is related to recorded data mainly referring to previous decades. The estimate of extremes over very long periods is heavily influenced by the small number of severe events that have occurred in recent history. The second concerns the spatial variability of recorded cyclone tracks. Because of the low frequency of cyclone occurrence, estimates of extremes made from a limited database can vary substantially over relatively small distances, even within a spatially limited region where it would be reasonable to expect homogeneous values. Statistical uncertainty grows exponentially, moving toward cyclonic peripheral areas affected very unfrequently by cyclone passages. The commonly used methods to account for this uncertainty are historical track shifting and deductive approaches, but a certain degree of subjectivity is intrinsic in these approaches. For this reason, the aim of this paper is to build a more proper statistical methodology to account for cyclone spatial variability on extreme evaluations. It is based on the Inverse First Order Reliability Method (IFORM) used to determine the bivariate exceedance probability of the wind speed/track distance for cyclonic events in any selected location. By means of numerical modelling, namely Simulating WAves Nearshore (SWAN), the proposed approach, through a logical sequence, predicts the cyclonic track characteristics for a certain return period which, at the selected location, induce the worst conditions in terms of maximum sea state during a cyclonic event. Compared with the historical methods, which consider the effects of all the cyclones occurring in the area, the method is able to account for spatial variability even within regions with a very low cyclonic risk. The methodology produces consistent results, avoiding spurious discontinuities between contiguous locations which presumably have very similar vulnerability to cyclones.

Journal of Physics: Conference Series, 2017

Corrosion is one of the principal causes of degradation to failure of marine structures. In pract... more Corrosion is one of the principal causes of degradation to failure of marine structures. In practice, localized corrosion is the most dangerous mode of attack and can result in serious failures, in particular in marine flowlines and inter-field lines, arousing serious concerns relatively to environmental impact. The progress in time of internal corrosion, the location along the route and across the pipe section, the development pattern and the depth of the loss of metal are a very complex issue: the most important factors are products characteristics, transport conditions over the operating lifespan, process fluid-dynamics, and pipeline geometrical configuration. Understanding which factors among them play the most important role is a key step to develop a model able to predict with enough accuracy the sections more exposed to risk of failure. Some factors play a crucial role at certain spatial scales while other factors at other scales. The Mutual Information Theory, intimately related to the concept of Shannon Entropy in Information theory, has been applied to detect the most important variables at each scale. Finally, the variables emerged from this analysis at each scale have been integrated in a predicting data driven model sensibly improving its performance.

arXiv (Cornell University), Nov 17, 2022

Strongly opinionated minorities can have a dramatic impact on the opinion dynamics of a large pop... more Strongly opinionated minorities can have a dramatic impact on the opinion dynamics of a large population. Two factions of inflexible minorities, polarised into two competing opinions, could lead the entire population to persistent indecision. Equivalently, populations can remain undecided when individuals sporadically change their opinion based on individual information rather than social information. Our analysis compares the cross-inhibition model with the voter model for decisions between equally good alternatives, and with the weighted voter model for decisions among alternatives characterised by different qualities. Here we show that cross-inhibition, contrary to the other two models, is a simple mechanism that allows the population to reach a stable majority for one alternative even in the presence of a relatively high amount of asocial behaviour. The results predicted by the mean-field models are confirmed by experiments with swarms of 100 locally interacting robots. This work suggests an answer to the longstanding question of why inhibitory signals are widespread in natural systems of collective decision making, and, at the same time, it proposes an efficient mechanism for designing resilient swarms of minimalistic robots.

IEEE Transactions on Wireless Communications, 2023

In this paper, a Universal Underwater Software Defined Modem (UniSDM) architecture is proposed th... more In this paper, a Universal Underwater Software Defined Modem (UniSDM) architecture is proposed that may operate in different modes (acoustic, magnetic induction, optical and RF), in order to utilize the advantages of each mode and accordingly satisfy the requirements of many latest use cases in underwater communication systems. A detailed description of the novel UniSDM architecture is presented first. The novelty of this architecture is its flexibility, i.e., allowing the designers to produce a device that may include any type of modes operating seamlessly and jointly by exchanging data, control and synchronization. Many challenges, including high system costs and coordination between different modes, are addressed in the paper. Moreover, numerical evaluation is conducted to assess the performance of the proposed UniSDM architecture. Finally, the performance evaluation shows that the utilization of the UniSDM allows to decrease the transmission latency and improve the energy efficiency, while maintaining high reliability and robustness in underwater communication systems.

OCEANS 2022, Hampton Roads

OCEANS 2022 - Chennai

Underwater manipulation is one of the most remarkable ongoing research subjects in robotics. Inte... more Underwater manipulation is one of the most remarkable ongoing research subjects in robotics. Intervention autonomous underwater vehicles (I-AUVs) not only have to cope with the technical challenges associated with traditional manipulation tasks but do so while currents and waves perturb the stability of the vehicle, and low-light, turbid water conditions impede perceiving the surroundings. Certainly, the dynamic nature and our limited understanding of the marine environment hinder the autonomous performance of underwater robot manipulation. This manuscript provides a discussion on previous research and the limiting factors that impose on the longenvisioned prospects of autonomous underwater manipulation to finally highlight research directions that have the potential to improve the autonomy capabilities of I-AUVs.

GLOBECOM 2022 - 2022 IEEE Global Communications Conference, Dec 4, 2022