24067 research outputs found
Sort by
AI-Powered Image To Recipe Generator
Get PDFIn recent years, Artificial Intelligence (AI) has transformed various aspects of food technology, including automated recipe generation. This paper introduces an AI-powered Image-to-Recipe Generator, a system that extracts ingredient details from an image and generates a structured recipe with cooking instructions. Utilizing computer vision (CNNs, Vision Transformers) for ingredient detection and natural language processing (NLP) for recipe formulation, this system enhances the culinary experience by enabling users to derive recipes from food images effortlessly. The model integrates deep learning-based image recognition, NLP-driven text generation, and a recommendation engine to provide
personalized recipes based on dietary preferences and nutritional constraints. This AI- powered solution bridges the gap between visual food identification and recipe creation,
offering a seamless and interactive way to explore cooking possibilities
AI-Driven mHealth Solutions for Regulating Screen Time in Young Children: A Systematic Review of Parental Interventions, Digital Well-Being, and HCI Approaches
No full textThe increasing prevalence of screen timing among people, especially young children, has become a concerning issue as this may impact directly or indirectly on children’s cognitive, emotional and physical development. To address this challenge, it is important to find out the underlying causes of this increasing screen timing and therefore, our research is focused on performing a systematic review of the available literature on screen time regulation, parental intervention strategies and HCI. In the review, our goal is to pinpoint significant research gaps concerning the personal and socioeconomic influences on digital consumption in young children by dissecting current research findings. Besides, we are going through the prior research to find the parental obstacles, digital well-being interventions and AI based solutions to regulate a healthy screen timing routine. Through this examination, we are focusing on the emerging need of adaptive, personalized and data-driven interventions which will help parents to manage their children’s unregulated screen timing. By using the gained insights, we aim to build a smartphone based mHealth solution which will leverage AI and Machine Learning algorithms. This tool will offer real time monitoring and personalized advice based on the digital patterns and the socio economic status of the parent child dyad. Thus, our initiative is aimed at closing the gap between the theoretical research and the practical application to support the optimal cognitive and emotional development of young children and, therefore, ensure the correct approach to the digital interaction
Invisible Scholars
No full textIn the United States, every year there are approximately 1.3 million students who experience homelessness, an individual who lacks a fixed, regular, and adequate nighttime residence. Homelessness among students has a significant impact on their academic performance, mental health, and behavior. Factors like higher absenteeism, constant movement, and lack of support can all contribute to decreased academic achievement. Ultimately, this leads to students having to repeat grades or not graduating high school, compromising their abilities to succeed in the future. In addition to the decrease in academic achievement, behavioral and mental challenges exacerbated by the stress of housing instability also affect the student’s development. In Macon, Georgia, there are currently five hundred students aged six to seventeen, who lack stable housing. These students have access to assistance through programs like the Mckinney Vento Homeless Assistance Act and Rapid Rehousing. Mckinney Vento Homeless Assistance Act helps students to have access to school and stay in school, while Rapid Rehousing aims to provide housing resources for students and their families. However, these programs have differing definitions of homelessness that lead to gaps in how many students are able to receive housing support. In downtown Macon, there is an urban renewal occurring aimed at bringing more life downtown. This, along with social and community services, influenced the decision to create a space for homeless students right in the heart of Macon. The aim of my thesis is to connect these students with resources and the community, as well as provide them with stability in their most developmental years
Preliminary Work on The Design and Development of Biomimetic Upper-Body Exoskeleton for Stroke Patients
No full textThis study presents the preliminary design and development of a biomimetic upper-body exoskeleton aimed at assisting individuals with impaired arm function. The exoskeleton consists of a three degrees of freedom shoulder design controlled by two pancake motors, and a biomimetic and tendon-driven elbow joint controlled by a servo motor. To optimize weight distribution, the servo motor is housed at the back of the waist, shifting the load toward the shoulder for improved user comfort. Motion replication is achieved using inertial measurement unit (IMU) sensors placed on the upper and lower arms to capture the healthy arm’s movements. The sensor data is wirelessly transmitted to a Raspberry Pi, where it is processed as quaternion inputs to represent the arm’s motion. The exoskeleton, controlled via CAN communication, is attached to the injured arm and actuated to mirror the movements of the healthy arm. The Raspberry Pi and motor controllers are integrated into a compact backpack worn by the user. Additionally, the MATLAB Simscape model is created to simulate the motion of the exoskeleton which provides a framework for refining future iterations of the compliant and biomimetic joints
The Translucent Fortress: Improving the Design of Rural Domestic Violence Shelters
Get PDFDomestic violence is a nationwide social plague that effects millions of women each year and has increased in severity since Covid-19 (2020-2022). In the state of Georgia, 37.4% of women report experiencing some sort of physical violence by an intimate partner (NCADV, 2020). Severity and frequency of domestic violence increases drastically if the victim is living in a rural location. Qualitative studies conducted in Iowa, North Carolina, and Virginia all report that on average, women living in rural areas report 150% more domestic violence experiences at a higher severity than their urban counterparts and must travel on average 300% further to access resources (Eastman et al, 2007; Peek-Asa et al, 2011). As well, rural women are nearly twice as likely to be turned away from services because the lack the necessary space, staffing, resources, and training to provide proper care for the victims (Peek-Asa et al, 2011). Also, the efficiency of existing shelters often falls short due to poor architectural design and lack of security (Lygum et al, 2013). Complications arise when approaching this issue due to factors such as the geographic isolation of victims, meaning less public visibility and greater distances between homes, religious pressure from the church and family members for women to stay with the abuser and maintain the family, and poverty levels creating difficulty of access to vehicles or public transportation (Eastman et al, 2007).
The building typology of a women’s shelter will be reimagined to better apply to a modern stage using qualitative research from an architect-conducted survey that has been compared to similar existing research. This new typology will be tested on a site in Homerville Georgia to address the psychological and spatial needs of victims, the most important of these needs include the feeling of being protected (secure architecture), a sense of community (social spaces), access to medical assistance (operating/examination rooms), and appropriate design to accommodate living spaces for both women and/or their children (Eastman et al, 2007). Construction of this project must be efficient and cost effective as to not diminish the already limited financial resources available to public services, but still secure enough to provide a safe space for rural victims of domestic violence.
Research for this thesis will be broken up into multiple different sections. To give the architect a better understanding of the victim’s condition and needs, surveys and interviews were conducted during the summer of 2024 at women’s shelters in rural Georgia with both the residents and shelter staff, and literature reviews were done to compare collected data with similar research. This provided an updated set of data to be used and compared with older information (pre Covid-19). To determine the site for the project, mapping was used to locate a site in Georgia with the least access to domestic violence crisis centers. Finally, construction methods and building design will be studied through precedents that share the scope of being low cost, remote, secure, and inhabitant-friendly design for the purpose of applying efficient design methods
SELA Officers and Executive Committee, SELA State Representatives
Get PDFCurrent SELA officers and representatives
New Suburbanism Learning From Suburbia
Get PDFWhy does current architecture not address suburban spaces? Suburban spaces are the bulk of what is seen in Atlanta Georgia; and are the most common lived space these spaces are known as Sprawl, this sprawl seen in Atlanta can be described as fragmented, contradictory, corporate, smooth, additive and ever expanding. Suburban spaces are also airy, convenient and it addresses the needs of our current societal needs. What happens when you embrace and intensify the junkspaces and non-places seen in suburbia? How do you live with the parking lot, the big box store and the strip mall? This thesis is about this design challenge, how do you embrace these common spaces and qualities through re-composition and layering of the non-place.
In Learning from Las vegas, Robert Venturi and Denise Scott Brown studied and used the design principles of the the Las Vegas Strip to create new forms and concepts in architecture. Along the same line, Rem Koolhaas has theorized about these spaces in Junkspace. Their work shows the result of embracing these common spaces. This research seeks to perform similar studies focusing on the spaces of Atlanta Georgia. This research will document the suburban typologies of Atlanta and perform precedent analysis of relevant projects. This research seeks to critically look at suburban architecture and understands its design principles to embrace and amplify the common spaces of Atlanta Georgia
Research on Brain Augmented Technology (BAT) and STEM-Peer Augmented Success & Support (STEM-PASS)
No full textThrough understanding and analyzing the electroencephalography (EEG) of attention and distraction experiments, it becomes possible to explore the effects that certain events related potentials (ERP) have on attention. Attention is essential throughout the entirety of individuals\u27 lives, while distraction plays an equally important role in how attention is affected. The research conducted in the VIP-BAT EEG lab aims to gather, analyze and interpret data, and draw conclusions on the study of Attention and Distraction. Using volunteer participants, these experiments focus on how brain waves correlate with results from gameplay and with external factors influencing the outcomes. When gathering data on attention and distraction, participants engage in several mini-games, with their environment changing each round. Specifically, in data capturing, EMOTIV hardware and software are used for wave capturing and MATLAB is used to process the data in graphical form to identify noticeable impacts. Referring to the experimental data helps create understanding of how the brain and body react to these events simultaneously. The results of this research will help in addressing a wide range of issues, from assisting a student with studying, to potentially saving a life. Authors: Dylan Andersen, Liam Begley, Ololade Caulker, Janice Ryoo, Dandre Williamso
AI in Beginning Design Pedagogy
No full textThis research investigates how Midjourney AI can be incorporated into the introductory pedagogy of first-year architecture students, a traditionally hands-on, analog phase of architectural education. The central research question asks: How can AI, particularly Midjourney, be used to support or enhance the design process for beginning architecture students without replacing foundational craft and design thinking skills? The purpose of this research is to assess both the opportunities and limitations of using AI in the earliest stages of architectural education, with a focus on ideation, visual exploration, and creative complexity.
As AI tools rapidly evolve and become increasingly embedded in professional architectural practice, understanding how students can engage with them early in their education is essential. This project aims to explore whether introducing AI at the foundational level can help students better understand contemporary design workflows, accelerate concept generation, and enhance their visual literacy—while avoiding overdependence on technology or loss of traditional design skills.
The methodology involved translating existing first-year architecture assignments into AI prompts, using ChatGPT to refine those prompts, and then generating imagery with Midjourney. Additionally, student work was incorporated as visual reference inputs to evaluate how AI could remix or expand upon physical models and drawings. Preliminary findings indicate that while Midjourney struggles to accurately follow assignment sheets or reproduce specific craft-based outputs, it excels at generating unexpected visual complexity and assisting with ideation. AI proved particularly effective when prompts were fed incrementally and combined with student work. Importantly, the AI’s inability to replicate polished student work alleviates concerns around academic integrity. This research highlights AI’s potential as a creative partner in the design studio and anticipates further development of structured prompt strategies to better align AI capabilities with architectural pedagogy
Characterizing the role of pmt-2 in C. elegans larval development
No full textNematode parasitic infections are becoming more prevalent in humans and animals. It is estimated that over 3 billion people have gotten at least one nematode infection in their lives, whether from hookworm, roundworm, or pinworm. Working on parasitic nematodes is difficult because they can be hard to culture in the lab. To circumvent this problem, we took the approach to working with Caenorhabditis elegans, a non-parasitic nematode that is easy to grow in culture. Studies have shown that nematodes synthesize phosphatidylcholine by a novel pathway which uses the enzymes PMT-1 and PMT-2. Our focus is to develop enzyme inhibitors of PMT-2, because the phosphatidylcholine biosynthetic pathway is crucial to the synthesis of membrane phospholipids. In particular, the PMT-2 enzyme is unique to nematodes, which should make it easier to target nematodes specifically. The pmt-2(ok2419) V/nT1[qIs51] strain is a homozygous lethal deletion mutation balanced by GFP-tagged translocation chromosome. Translocation balancers are difficult to work with because of aneuploidy, which is when there is an abnormal number of chromosomes in a cell. To facilitate our characterization of the pmt-2(ok2419) mutation, we set out to build a simple linked inversion balancer. Ten wildtype males were crossed with tmC16 homozygous mutants, then the resulting heterozygous F1 males crossed with oxTi396[eft-3p::mCherry] hermaphrodites. This cross was transferred daily to new plates to not have any self-fertilization with progeny in newer generations. These two reporters together are expected to have a mCherry marker in the pharynx and broad red nuclear fluorescence in body wall muscles and other tissues. Our preliminary data indicates that PMT-2 mutants may fail to molt out of the first larval stage. Work is ongoing to better characterize the role PMT-2 in larval development