-
Deconstructing Bias: A Multifaceted Framework for Diagnosing Cultural and Compositional Inequities in Text-to-Image Generative Models
Authors:
Muna Numan Said,
Aarib Zaidi,
Rabia Usman,
Sonia Okon,
Praneeth Medepalli,
Kevin Zhu,
Vasu Sharma,
Sean O'Brien
Abstract:
The transformative potential of text-to-image (T2I) models hinges on their ability to synthesize culturally diverse, photorealistic images from textual prompts. However, these models often perpetuate cultural biases embedded within their training data, leading to systemic misrepresentations. This paper benchmarks the Component Inclusion Score (CIS), a metric designed to evaluate the fidelity of im…
▽ More
The transformative potential of text-to-image (T2I) models hinges on their ability to synthesize culturally diverse, photorealistic images from textual prompts. However, these models often perpetuate cultural biases embedded within their training data, leading to systemic misrepresentations. This paper benchmarks the Component Inclusion Score (CIS), a metric designed to evaluate the fidelity of image generation across cultural contexts. Through extensive analysis involving 2,400 images, we quantify biases in terms of compositional fragility and contextual misalignment, revealing significant performance gaps between Western and non-Western cultural prompts. Our findings underscore the impact of data imbalance, attention entropy, and embedding superposition on model fairness. By benchmarking models like Stable Diffusion with CIS, we provide insights into architectural and data-centric interventions for enhancing cultural inclusivity in AI-generated imagery. This work advances the field by offering a comprehensive tool for diagnosing and mitigating biases in T2I generation, advocating for more equitable AI systems.
△ Less
Submitted 5 April, 2025;
originally announced May 2025.
-
Identifying Human Indoor Daily Life Behavior employing Thermal Sensor Arrays (TSAs)
Authors:
Dina E. Abdelaleem,
Hassan M. Ahmed,
M. Sami Soliman,
Tarek M. Said
Abstract:
Daily activity monitoring systems used in households provide vital information for health status, particularly with aging residents. Multiple approaches have been introduced to achieve such goals, typically obtrusive and non-obtrusive. Amongst the obtrusive approaches are the wearable devices, and among the non-obtrusive approaches are the movement detection systems, including motion sensors and t…
▽ More
Daily activity monitoring systems used in households provide vital information for health status, particularly with aging residents. Multiple approaches have been introduced to achieve such goals, typically obtrusive and non-obtrusive. Amongst the obtrusive approaches are the wearable devices, and among the non-obtrusive approaches are the movement detection systems, including motion sensors and thermal sensor arrays (TSAs). TSA systems are advantageous when preserving a person's privacy and picking his precise spatial location. In this study, human daily living activities were monitored day and night, constructing the corresponding activity time series and spatial probability distribution and employing a TSA system. The monitored activities are classified into two categories: sleeping and daily activity. Results showed the possibility of distinguishing between classes regardless of day and night. The obtained sleep activity duration was compared with previous research using the same raw data. Results showed that the duration of sleep activity, on average, was 9 hours/day, and daily life activity was 7 hours/day. The person's spatial probability distribution was determined using the bivariate distribution for the monitored location. In conclusion, the results showed that sleeping activity was dominant. Our study showed that TSAs were the optimum choice when monitoring human activity. Our proposed approach tackled limitations encountered by previous human activity monitoring systems, such as preserving human privacy while knowing his precise spatial location.
△ Less
Submitted 12 September, 2024;
originally announced September 2024.
-
Deep learning in computed tomography pulmonary angiography imaging: a dual-pronged approach for pulmonary embolism detection
Authors:
Fabiha Bushra,
Muhammad E. H. Chowdhury,
Rusab Sarmun,
Saidul Kabir,
Menatalla Said,
Sohaib Bassam Zoghoul,
Adam Mushtak,
Israa Al-Hashimi,
Abdulrahman Alqahtani,
Anwarul Hasan
Abstract:
The increasing reliance on Computed Tomography Pulmonary Angiography (CTPA) for Pulmonary Embolism (PE) diagnosis presents challenges and a pressing need for improved diagnostic solutions. The primary objective of this study is to leverage deep learning techniques to enhance the Computer Assisted Diagnosis (CAD) of PE. With this aim, we propose a classifier-guided detection approach that effective…
▽ More
The increasing reliance on Computed Tomography Pulmonary Angiography (CTPA) for Pulmonary Embolism (PE) diagnosis presents challenges and a pressing need for improved diagnostic solutions. The primary objective of this study is to leverage deep learning techniques to enhance the Computer Assisted Diagnosis (CAD) of PE. With this aim, we propose a classifier-guided detection approach that effectively leverages the classifier's probabilistic inference to direct the detection predictions, marking a novel contribution in the domain of automated PE diagnosis. Our classification system includes an Attention-Guided Convolutional Neural Network (AG-CNN) that uses local context by employing an attention mechanism. This approach emulates a human expert's attention by looking at both global appearances and local lesion regions before making a decision. The classifier demonstrates robust performance on the FUMPE dataset, achieving an AUROC of 0.927, sensitivity of 0.862, specificity of 0.879, and an F1-score of 0.805 with the Inception-v3 backbone architecture. Moreover, AG-CNN outperforms the baseline DenseNet-121 model, achieving an 8.1% AUROC gain. While previous research has mostly focused on finding PE in the main arteries, our use of cutting-edge object detection models and ensembling techniques greatly improves the accuracy of detecting small embolisms in the peripheral arteries. Finally, our proposed classifier-guided detection approach further refines the detection metrics, contributing new state-of-the-art to the community: mAP$_{50}$, sensitivity, and F1-score of 0.846, 0.901, and 0.779, respectively, outperforming the former benchmark with a significant 3.7% improvement in mAP$_{50}$. Our research aims to elevate PE patient care by integrating AI solutions into clinical workflows, highlighting the potential of human-AI collaboration in medical diagnostics.
△ Less
Submitted 5 January, 2024; v1 submitted 9 November, 2023;
originally announced November 2023.
-
AI Models Close to your Chest: Robust Federated Learning Strategies for Multi-site CT
Authors:
Edward H. Lee,
Brendan Kelly,
Emre Altinmakas,
Hakan Dogan,
Maryam Mohammadzadeh,
Errol Colak,
Steve Fu,
Olivia Choudhury,
Ujjwal Ratan,
Felipe Kitamura,
Hernan Chaves,
Jimmy Zheng,
Mourad Said,
Eduardo Reis,
Jaekwang Lim,
Patricia Yokoo,
Courtney Mitchell,
Golnaz Houshmand,
Marzyeh Ghassemi,
Ronan Killeen,
Wendy Qiu,
Joel Hayden,
Farnaz Rafiee,
Chad Klochko,
Nicholas Bevins
, et al. (5 additional authors not shown)
Abstract:
While it is well known that population differences from genetics, sex, race, and environmental factors contribute to disease, AI studies in medicine have largely focused on locoregional patient cohorts with less diverse data sources. Such limitation stems from barriers to large-scale data share and ethical concerns over data privacy. Federated learning (FL) is one potential pathway for AI developm…
▽ More
While it is well known that population differences from genetics, sex, race, and environmental factors contribute to disease, AI studies in medicine have largely focused on locoregional patient cohorts with less diverse data sources. Such limitation stems from barriers to large-scale data share and ethical concerns over data privacy. Federated learning (FL) is one potential pathway for AI development that enables learning across hospitals without data share. In this study, we show the results of various FL strategies on one of the largest and most diverse COVID-19 chest CT datasets: 21 participating hospitals across five continents that comprise >10,000 patients with >1 million images. We also propose an FL strategy that leverages synthetically generated data to overcome class and size imbalances. We also describe the sources of data heterogeneity in the context of FL, and show how even among the correctly labeled populations, disparities can arise due to these biases.
△ Less
Submitted 13 April, 2023; v1 submitted 23 March, 2023;
originally announced March 2023.
-
Shifting Mobility Behaviors in Unprecedented Times: Intentions to Use On-demand Ride Services During the COVID-19 Pandemic
Authors:
Maher Said,
Jason Soria,
Amanda Stathopoulos
Abstract:
The spread of COVID-19 has been a major disruptive force in people's everyday lives and mobility behavior. The demand for on-demand ride services, such as taxis and ridehailing has been specifically impacted, given both restrictions in service operations and user's concerns about virus transmission in shared vehicles. During the pandemic, demand for these modes have decreased by as much as 80%. Th…
▽ More
The spread of COVID-19 has been a major disruptive force in people's everyday lives and mobility behavior. The demand for on-demand ride services, such as taxis and ridehailing has been specifically impacted, given both restrictions in service operations and user's concerns about virus transmission in shared vehicles. During the pandemic, demand for these modes have decreased by as much as 80%. This study examines intentions to use on-demand ride services in a period of drastic changes in lifestyles and daily routines coupled with unprecedented mobility reductions. Specifically, we examine the determinants for the shift of intentions to use these on-demand modes of travel in the early stages of the pandemic. Using data from a survey disseminated in June 2020 to 700 respondents from contiguous United States, ordinal regression modeling is applied to analyze the shift in consideration. The results indicate that political orientation and health-related experiences during the pandemic are significant sources of variation for individual changes in intentions to use ridehailing. Additionally, characteristics such as age and income result in consideration shifts that contradict the typical ridership profiles found in the ridehailing literature. Specifically, on-demand ride consideration decreases as a function of age and income. Moreover, transit-users are more willing to consider on-demand rides than private vehicle users, suggesting that shared vehicle modes have a similar risk-profile. We discuss the role of on-demand ride services in the pandemic era, and the need to investigate political orientation and evolving pandemic experiences to pinpoint their role in future mobility systems.
△ Less
Submitted 4 August, 2021;
originally announced August 2021.
-
Accelerating the Adoption of Disruptive Technologies: The Impact of COVID-19 on Intentions to Use Autonomous Vehicles
Authors:
Maher Said,
Emma R. Zajdela,
Amanda Stathopoulos
Abstract:
One of the most notable global transportation trends is the accelerated pace of development in vehicle automation technologies. Uncertainty surrounds the future of automated mobility as there is no clear consensus on potential adoption patterns, ownership versus shared use status and travel impacts. Adding to this uncertainty is the impact of the COVID-19 pandemic that has triggered profound chang…
▽ More
One of the most notable global transportation trends is the accelerated pace of development in vehicle automation technologies. Uncertainty surrounds the future of automated mobility as there is no clear consensus on potential adoption patterns, ownership versus shared use status and travel impacts. Adding to this uncertainty is the impact of the COVID-19 pandemic that has triggered profound changes in mobility behaviors as well as accelerated adoption of new technologies at an unprecedented rate. Accordingly, this study examines the impact of the COVID-19 pandemic on willingness to adopt the emerging technology of autonomous vehicles (AVs). Using data from a survey disseminated in June 2020 to 700 respondents in the United States, we perform a difference-in-difference regression to analyze the shift in willingness to use autonomous vehicles as part of a shared fleet before and during the pandemic. The results reveal that the COVID-19 pandemic has a positive and highly significant impact on the consideration of using autonomous vehicles. This shift is present regardless of tech-savviness, gender or urban/rural household location. Individuals who are younger, left-leaning and frequent users of shared modes of travel are expected to become more likely to use autonomous vehicles once offered. Understanding the effects of these attributes on the increase in consideration of AVs is important for policy making, as these effects provide a guide to predicting adoption of autonomous vehicles - once available - and to identify segments of the population likely to be more resistant to adopting AVs.
△ Less
Submitted 1 December, 2021; v1 submitted 3 August, 2021;
originally announced August 2021.
-
Interdependence in active mobility adoption: Joint modelling and motivational spill-over in walking, cycling and bike-sharing
Authors:
M Said,
A Biehl,
A Stathopoulos
Abstract:
Active mobility offers an array of physical, emotional, and social wellbeing benefits. However, with the proliferation of the sharing economy, new nonmotorized means of transport are entering the fold, complementing some existing mobility options while competing with others. The purpose of this research study is to investigate the adoption of three active travel modes; namely walking, cycling and…
▽ More
Active mobility offers an array of physical, emotional, and social wellbeing benefits. However, with the proliferation of the sharing economy, new nonmotorized means of transport are entering the fold, complementing some existing mobility options while competing with others. The purpose of this research study is to investigate the adoption of three active travel modes; namely walking, cycling and bikesharing, in a joint modeling framework. The analysis is based on an adaptation of the stages of change framework, which originates from the health behavior sciences. Multivariate ordered probit modeling drawing on U.S. survey data provides well-needed insights into individuals preparedness to adopt multiple active modes as a function of personal, neighborhood and psychosocial factors. The research suggests three important findings. 1) The joint model structure confirms interdependence among different active mobility choices. The strongest complementarity is found for walking and cycling adoption. 2) Each mode has a distinctive adoption path with either three or four separate stages. We discuss the implications of derived stage-thresholds and plot adoption contours for selected scenarios. 3) Psychological and neighborhood variables generate more coupling among active modes than individual and household factors. Specifically, identifying strongly with active mobility aspirations, experiences with multimodal travel, possessing better navigational skills, along with supportive local community norms are the factors that appear to drive the joint adoption decisions. This study contributes to the understanding of how decisions within the same functional domain are related and help to design policies that promote active mobility by identifying positive spillovers and joint determinants.
△ Less
Submitted 29 October, 2020; v1 submitted 24 June, 2020;
originally announced June 2020.
-
Optimization Based Solutions for Control and State Estimation in Non-holonomic Mobile Robots: Stability, Distributed Control, and Relative Localization
Authors:
Mohamed W. Mehrez Said
Abstract:
Interest in designing, manufacturing, and using autonomous robots has been rapidly growing during the most recent decade. The main motivation for this interest is the wide range of potential applications these autonomous systems can serve in. The applications include, but are not limited to, area coverage, patrolling missions, perimeter surveillance, search and rescue missions, and situational awa…
▽ More
Interest in designing, manufacturing, and using autonomous robots has been rapidly growing during the most recent decade. The main motivation for this interest is the wide range of potential applications these autonomous systems can serve in. The applications include, but are not limited to, area coverage, patrolling missions, perimeter surveillance, search and rescue missions, and situational awareness. In this thesis, the area of control and state estimation in non-holonomic mobile robots is tackled. Herein, optimization based solutions for control and state estimation are designed, analyzed, and implemented to such systems. One of the main motivations for considering such solutions is their ability of handling constrained and nonlinear systems such as non-holonomic mobile robots. Moreover, the recent developments in dynamic optimization algorithms as well as in computer processing facilitated the real-time implementation of such optimization based methods in embedded computer systems.
△ Less
Submitted 15 March, 2018;
originally announced March 2018.
-
On the Maximal Shortest Path in a Connected Component in V2V
Authors:
Michel Marot,
Adel Mounir Saïd,
Hossam Afifi
Abstract:
In this work, a VANET (Vehicular Ad-hoc NETwork) is considered to operate on a simple lane, without infrastructure. The arrivals of vehicles are assumed to be general with any traffic and speed assumptions. The vehicles communicate through the shortest path. In this paper, we study the probability distribution of the number of hops on the maximal shortest path in a connected component of vehicles.…
▽ More
In this work, a VANET (Vehicular Ad-hoc NETwork) is considered to operate on a simple lane, without infrastructure. The arrivals of vehicles are assumed to be general with any traffic and speed assumptions. The vehicles communicate through the shortest path. In this paper, we study the probability distribution of the number of hops on the maximal shortest path in a connected component of vehicles. The general formulation is given for any assumption of road traffic. Then, it is applied to calculate the z-transform of this distribution for medium and dense networks in the Poisson case. Our model is validated with the Madrid road traces of the Universitat Politècnica de Catalunya. These results may be useful for example when evaluating diffusion protocols through the shortest path in a VANET, where not only the mean but also the other moments are needed to derive accurate results.
△ Less
Submitted 8 October, 2015;
originally announced October 2015.
-
A Hybrid Model to Extend Vehicular Intercommunication V2V through D2D Architecture
Authors:
Emad Abd-Elrahman,
Adel Mounir Said,
Thouraya Toukabri,
Hossam Afifi,
Michel Marot
Abstract:
In the recent years, many solutions for Vehicle to Vehicle (V2V) communication were proposed to overcome failure problems (also known as dead ends). This paper proposes a novel framework for V2V failure recovery using Device-to-Device (D2D) communications. Based on the unified Intelligent Transportation Systems (ITS) architecture, LTE-based D2D mechanisms can improve V2V dead ends failure recovery…
▽ More
In the recent years, many solutions for Vehicle to Vehicle (V2V) communication were proposed to overcome failure problems (also known as dead ends). This paper proposes a novel framework for V2V failure recovery using Device-to-Device (D2D) communications. Based on the unified Intelligent Transportation Systems (ITS) architecture, LTE-based D2D mechanisms can improve V2V dead ends failure recovery delays. This new paradigm of hybrid V2V-D2D communications overcomes the limitations of traditional V2V routing techniques. According to NS2 simulation results, the proposed hybrid model decreases the end to end delay (E2E) of messages delivery. A complete comparison of different D2D use cases (best & worst scenarios) is presented to show the enhancements brought by our solution compared to traditional V2V techniques.
△ Less
Submitted 20 February, 2015;
originally announced February 2015.
-
Assisting V2V failure recovery using Device-to-Device Communications
Authors:
Emad Abd-Elrahman,
Adel Mounir Said,
Thouraya Toukabri,
Hossam Afifi,
Michel Marot
Abstract:
This paper aims to propose a new solution for failure recovery (dead-ends) in Vehicle to Vehicle (V2V) communications through LTE-assisted Device-to-Device communications (D2D). Based on the enhanced networking capabilities offered by Intelligent Transportation Systems (ITS) architecture, our solution can efficiently assist V2V communications in failure recovery situations. We also derive an analy…
▽ More
This paper aims to propose a new solution for failure recovery (dead-ends) in Vehicle to Vehicle (V2V) communications through LTE-assisted Device-to-Device communications (D2D). Based on the enhanced networking capabilities offered by Intelligent Transportation Systems (ITS) architecture, our solution can efficiently assist V2V communications in failure recovery situations. We also derive an analytical model to evaluate generic V2V routing recovery failures. Moreover, the proposed hybrid model is simulated and compared to the generic model under different constrains of worst and best cases of D2D discovery and communication. According to our comparison and simulation results, the hybrid model decreases the delay for alarm message propagation to the destination (typically the Traffic Control Center TCC) through the Road Side Unit (RSU)
△ Less
Submitted 5 February, 2015;
originally announced February 2015.
-
Independent Component Analysis for Filtering Airwaves in Seabed Logging Application
Authors:
Adeel Ansari,
Afza Bt Shafie,
Abas B Md Said,
Seema Ansari
Abstract:
Marine controlled source electromagnetic (CSEM) sensing method used for the detection of hydrocarbons based reservoirs in seabed logging application does not perform well due to the presence of the airwaves (or sea-surface). These airwaves interfere with the signal that comes from the subsurface seafloor and also tend to dominate in the receiver response at larger offsets. The task is to identify…
▽ More
Marine controlled source electromagnetic (CSEM) sensing method used for the detection of hydrocarbons based reservoirs in seabed logging application does not perform well due to the presence of the airwaves (or sea-surface). These airwaves interfere with the signal that comes from the subsurface seafloor and also tend to dominate in the receiver response at larger offsets. The task is to identify these air waves and the way they interact, and to filter them out. In this paper, a popular method for counteracting with the above stated problem scenario is Independent Component Analysis (ICA). Independent component analysis (ICA) is a statistical method for transforming an observed multidimensional or multivariate dataset into its constituent components (sources) that are statistically as independent from each other as possible. ICA-type de-convolution algorithm that is FASTICA is considered for mixed signals de-convolution and considered convenient depending upon the nature of the source and noise model. The results from the FASTICA algorithm are shown and evaluated. In this paper, we present the FASTICA algorithm for the seabed logging application.
△ Less
Submitted 4 March, 2013;
originally announced March 2013.
-
Personality wireless sensor networks (PWSNs)
Authors:
Muhammad Imran,
Halabi B. Hasbullah,
Abas Md Said
Abstract:
In recent years, WSNs are garnering lot of interest from research community because of their unique characteristics and potential for enormous range of applications. Envision for new class of applications are being emerged such as human augmentation, enhancing social interaction etc. Misunderstanding or misinterpretation of behaviors from individuals leads to social conflicts. There are various th…
▽ More
In recent years, WSNs are garnering lot of interest from research community because of their unique characteristics and potential for enormous range of applications. Envision for new class of applications are being emerged such as human augmentation, enhancing social interaction etc. Misunderstanding or misinterpretation of behaviors from individuals leads to social conflicts. There are various theories that classify people into different personality types. Most of the existing theories rely on questionnaires, which is highly unreliable. Anyone can lead such theories in practice to incorrect classification intentionally or unintentionally. The objective of this research is to investigate existing solutions and propose a basic infrastructure for an automated context-aware psychological classification based on different parameters. The idea is to use wearable sensors to sense and measure various human body parameters (i.e. body temperature, blood pressure, perspiration, brain impulses etc) that coerce human psychological condition. The data collected from these parameters is transformed in to information, to determine personality type, mood and psychological condition of interacting parties. This information is shared among counterparts to better understand each other in order to avoid potential conflicting situations. We believe that it will help peoples understand each other, improve their quality of life and minimize possible conflicting situations.
△ Less
Submitted 21 December, 2012;
originally announced December 2012.
