Browsing by Subject "artificial intelligence"

Now showing 1 - 9 of 9
  • Thumbnail Image
    Item
    Artificial Intelligence Governance: A Comparative Analysis of China, the European Union, and the United States
    (2022-05) Dixon, Ren Bin Lee
    Artificial Intelligence (AI) has become increasingly more ubiquitous and deployed across many sectors and industries. While the technology is expected to bring transformative changes to society, there has been a growing urgency to establish robust governance frameworks to mitigate the issues and risks attendant with its deployment. A representative governance initiative was selected from China, the European Union, and the United States — as the three leading global AI regimes at present — to conduct a comparative analysis on their approaches. These policy recommendations were developed to address fundamental AI principles that had been identified and distilled from a corpus of over ninety AI governance initiatives published by the academia, private and public sectors, and multi-stakeholder groups. AI principles were chosen as the standard for policy analysis in this paper because they have been established — in the field of AI governance — as well-researched guidelines that can be used as the foundation for developing AI governance frameworks.
  • Thumbnail Image
    Item
    Enhancing mammal conservation in multi-functional landscapes using artificial intelligence, joint species distribution modeling and ecological experimentation
    (2022-12) Velez Gomez, Juliana
    Poaching and livestock production threaten wildlife and its habitat, requiring strategies to manage human-dominated landscapes to sustain conservation objectives. To better understand the spatiotemporal distribution of wildlife and its response to disturbance factors (i.e., poaching and cattle), I deployed camera traps (CTs) and automated acoustic recording units (ARUs) on cattle ranches in the Colombian Orinoquía region. Data collection resulted in the challenge of processing “Big Data,” comprising a total of 824,883 images and 3,491,528 audio files (25,584 hours of recordings). In Chapter 1, I evaluated artificial intelligence platforms built for processing CT data and developed an open-source GitBook that illustrates the use and evaluation of model performance of each of these platforms. In Chapter 2, I used CT data to detect wildlife and trained machine learning algorithms for detecting cattle and poaching activity from the ARU data. To quantify co-occurrence patterns of poachers, cattle, and wildlife, I analyzed these data using joint species distribution models, finding that co-occurrence patterns between disturbance and wild ungulates were dependent on the data-collection method (i.e., whether CTs or ARUs were used to detect disturbance). Lastly, in Chapter 3, I conducted a cattle exclusion experiment to evaluate the effectiveness of fencing for reducing forest use and habitat degradation by cattle and improving resource availability for wildlife. Collectively, these efforts will guide management in multi-functional landscapes by identifying spatial co-occurrence patterns between wildlife and disturbance factors and by scaling up evidence-based interventions to optimize the use of remaining habitat by wildlife.
  • Thumbnail Image
    Item
    Fusion of Knowledge: Enhancing AI Reasoning through Language Models and Knowledge Graphs
    (2024-06) Mavromatis, Konstantinos
    Large Language Models (LLMs) and Knowledge Graphs (KGs) have rapidly emerged as important areas in Artificial Intelligence (AI). LLMs leverage vast amounts of unstructured text to understand and generate natural language. KGs are relational graphs that encode domain expertise and knowledge into explicit semantics. A desideratum of AI is the ability to reason and draw inferences in a rational, sensible way. The present dissertation addresses the following question: How can LLMs and KGs enhance AI reasoning? The core idea of this dissertation is to leverage LLMs as a foundation for understanding and processing natural language, while utilizing KGs to access accurate and domain-specific knowledge. We present our contributions in advancing the capabilities of AI systems with respect to the following dimensions. (1) Faithfulness: We introduce a novel KG retrieval method (GNN-RAG) for grounding the LLM reasoning on multi-hop KG facts, alleviating LLM hallucinations when answering complex questions. (2) Effectiveness: We design a powerful graph model (ReaRev) for improved reasoning over KGs on knowledge-intensive tasks, such as Question Answering. (3) Temporal Reasoning: We propose TempoQR, a method that leverages Temporal KGs and allows LMs to handle questions with temporal constraints. (4) Efficiency: We develop a graph-aware distillation framework (GRAD), in which the LM learns to utilize useful graph information, while being efficient at inference. (5) Robustness: We present SemPool, a simple graph pooling method that offers robustness when critical information is missing from the KG.
  • Thumbnail Image
    Item
    A High-Precision Bioelectric Neural Interface Toward Human-Machine Symbiosis
    (2021-01) Nguyen, Anh Tuan
    Objective. A symbiosis of human intelligence and artificial intelligence (AI) cannot be achieved without establishing an intuitive, bidirectional, and high-bandwidth information conduit between the minds and machines. Approach. Here we focus on developing high-precision bioelectronics underlying a new class of bioelectric neural interfaces that could bring us one step closer to this feat. We pioneer new circuit techniques, including frequency shaping (FS), redundant sensing (RS), RS-based super-resolution, and redundant crossfire (RXF), to enhance the effective resolution of neural recording and stimulation. These fundamentals allow the implementation of a series of fully-integrated microchips called Neuronix capable of acquiring low-noise neural signals and delivering high-precision electrical microstimulation. The Neuronix chips are incorporated to create miniaturized neuromodulation devices, including the Scorpius system, to enable bidirectional communications with neural circuits. Results. In a clinical study with human amputees, the Scorpius system helps establish a peripheral nerve interface that allows deep learning-based AI models to read and decode the patients' intents of moving individual fingers. Our analysis of acquired electroneurography (ENG) signals demonstrates this robust nerve interface has a sufficient information capacity to enable real-time control of a multi-degree-of-freedom (DOF) neuroprosthetic hand with near-natural dexterity and intuitiveness while simultaneously delivering somatosensory feedback. Significance. Our study layouts the principled foundation toward not only a dexterous control strategy for advanced neuroprostheses but also an intuitive conduit for connecting the human minds and machines. This opens up possibilities for many biomedical applications and manifests the basis of the future human-machine symbiosis.
  • Thumbnail Image
    Item
    Human Guidance Behavior Decomposition and Modeling
    (2017-12) Feit, Andrew
    Trained humans are capable of high performance, adaptable, and robust first-person dynamic motion guidance behavior. This behavior is exhibited in a wide variety of activities such as driving, piloting aircraft, skiing, biking, and many others. Human performance in such activities far exceeds the current capability of autonomous systems in terms of adaptability to new tasks, real-time motion planning, robustness, and trading safety for performance. The present work investigates the structure of human dynamic motion guidance that enables these performance qualities. This work uses a first-person experimental framework that presents a driving task to the subject, measuring control inputs, vehicle motion, and operator visual gaze movement. The resulting data is decomposed into subspace segment clusters that form primitive elements of action-perception interactive behavior. Subspace clusters are defined by both agent-environment system dynamic constraints and operator control strategies. A key contribution of this work is to define transitions between subspace cluster segments, or subgoals, as points where the set of active constraints, either system or operator defined, changes. This definition provides necessary conditions to determine transition points for a given task-environment scenario that allow a solution trajectory to be planned from known behavior elements. In addition, human gaze behavior during this task contains predictive behavior elements, indicating that the identified control modes are internally modeled. Based on these ideas, a generative, autonomous guidance framework is introduced that efficiently generates optimal dynamic motion behavior in new tasks. The new subgoal planning algorithm is shown to generate solutions to certain tasks more quickly than existing approaches currently used in robotics.
  • Thumbnail Image
    Item
    Machine Learning Techniques for Time Series Regression in Unmonitored Environmental Systems
    (2023-04) Willard, Jared
    This thesis provides a computer science audience with a review of machine learningtechniques for modeling time series in unmonitored environmental systems with no available target data that have been published in recent years, and further includes three distinct research efforts applying these methods to real-world water resources prediction scenarios. Additionally, we identify several open questions for time series prediction in unmonitored sites that include incorporating dynamic inputs and site characteristics, mechanistic understanding, and explainable AI techniques in modern machine learning frameworks. This is motivated by the current state of environmental time series modeling seeing a vast increase in applications of various machine learning models, in particular deep learning models built using the growing availability of high performance computing resources. It remains difficult to predict environmental variables for which observations are concentrated in a minority of locations and most locations remain unmonitored, and although many machine learning-based approaches have been developed, there is often a lack of comparison between them. The increased attention to environmental prediction topics such as disaster response, water resources management, and climate change reveal a need to compare these approaches, and understand when and where they should be applied in unmonitored environmental prediction scenarios.
  • Thumbnail Image
    Item
    Oral History Interview with Allen R. Hanson
    (Charles Babbage Institute, 2022-02) Hanson, Allen
    This interview was conducted by CBI for CS&E in conjunction with the 50th Anniversary of the University of Minnesota Computer Science Department (now Computer Science and Engineering, CS&E). Professor Hanson briefly discusses his early education and interests through his graduate education completing his doctorate at Cornell (dissertation was on games and prediction problems). Most of the interview focuses on his career and he was one of the early faculty members of the newly formed Computer Science Department at the University of Minnesota. He discusses the early department, interaction, and teaching, and research. His research focused heavily on vision and computing, pattern recognition, and AI. He partnered on early research with University of Massachusetts Amherst’s Ed Riseman and later left University of Minnesota to join the CS faculty of UMass and lead the lab in this collaboration. Among other topics, he outlines his evolving research, applications in medicine, autonomous vehicles and other areas, as well as reflects on a range of issues on research funding, and computing and society. Finally, he briefly discusses Applied Imaging, Dataviews and concurrent enterprises he led/helped to lead.
  • Thumbnail Image
    Item
    Physics-Driven Deep Learning Techniques for High-Resolution MRI
    (2023-05) Demirel, Omer
    Magnetic resonance imaging (MRI) is a non-invasive diagnostic tool used in clinics to evaluate the functional properties of the human body with superior soft-tissue contrast. Scan duration is a major issue in MRI that requires trade-offs between signal-to-noise ratio (SNR), spatio-temporal resolution, and coverage leading to numerous challenges. The need for faster MRI acquisition is particularly important in cardiac imaging and functional MRI (fMRI), where improved spatio-temporal resolution is essential for better coverage and evaluation. To tackle these issues, accelerated MRI techniques have been developed, such as parallel imaging, simultaneous multi-slice (SMS) imaging and compressed sensing (CS). Although advanced image reconstruction techniques are applied to reduce scan time while maintaining high-quality images, these techniques are limited in certain ways. Hand-crafted sparsity assumptions, blocky artifacts due to reconstruction errors, time-consuming parameter tuning, and long reconstruction times due to the iterative nature of the algorithms are the main limitations of CS. Recently, physics-guided deep learning (PG-DL) or physics-driven deep learning (PD-DL) reconstruction has gained immense interest in fast MRI. PD-DL is particularly useful because it combines the benefits of MRI physics with advanced neural network-based regularization techniques. On the other hand, PD-DL has already shown improved image quality compared to parallel imaging and CS and has led to unprecedented acceleration rates. Yet, PD-DL has its own limitations some of which are, limited training raw-data availability, overregularization or artificially hallucinated image features, generalizability issues across different SNRs, and sensitivity to noise. In this thesis, novel reconstruction methods were introduced to address these challenges using parallel imaging, cutting-edge SMS techniques, and state-of-the-art PD-DL reconstruction. First, we introduced an SMS reconstruction technique that was applied to cardiac MRI (CMR) to achieve faster heart coverage without compromising the image quality. This method addressed noise amplification and inter-slice leakage problems in accelerated SMS imaging using an extended k-space approach where SMS acceleration was characterized as a uniform sub-sampling in the readout direction. Second, we proposed to encode signal intensity variations across image phases into the forward operator of the MRI inverse problem without altering coil sensitivities to tackle the generalizability issue across different SNRs/contrast. This led to a more uniform contrast across the image series, which in turn facilitated generalizability for PD-DL methods. Third, we proposed to use a subject-specific self-supervised physics-guided deep learning reconstruction that exploits spatio-temporal correlations by using a 3D convolutional neural network. This network was trained on a subject of interest without a database to overcome the challenging database learning process of cardiac motion patterns for free-breathing CMR. Fourth, we extended a self-supervised PG-DL reconstruction to 20-fold accelerated 7T fMRI to show functional precision and temporal effects in the subsequent fMRI analysis were not altered by deep learning reconstruction leading to ultra-high acceleration rates with SMS and in-plane acceleration. Lastly, we re-envisioned the conventional fMRI computational imaging pipeline. Instead of performing reconstruction followed by denoising, we achieved improved image quality by employing PG-DL reconstruction after denoising the raw k-space leading to a synergistic combination of thermal noise suppression followed by deep learning reconstruction which leveraged the best of both worlds.
  • Thumbnail Image
    Item
    Siri, Stereotypes, and the Mechanics of Sexism
    (Western Libraries, 2022-12-21) Elder, Alexis M
    Feminized AIs designed for in-home verbal assistance are often subjected to gendered verbal abuse by their users. I survey a variety of features contributing to this phenomenon—from financial incentives for businesses to build products likely to provoke gendered abuse, to the impact of such behavior on household members—and identify a potential worry for attempts to criticize the phenomenon; while critics may be tempted to argue that engaging in gendered abuse of AI increases the chances that one will direct this abuse toward human beings, the recent history of attempts to connect video game violence to real-world aggression suggests that things may not be so simple. I turn to Confucian discussions of the role of ritualized social interactions both to better understand the roots of the problem and to investigate potential strategies for improvement, given a complex interplay between designers and device users. I argue that designers must grapple with the entrenched sexism in our society, at the expense of “smooth” and “seamless” user interfaces, in order to intentionally disrupt entrenched but harmful patterns of interaction, but that doing so is both consistent with and recommended by Confucian accounts of social rituals.