Diffusion models excel at generating photo-realistic images but come with significant computational costs in both training and sampling. While various techniques address these computational challenges, a less-explored issue is designing an efficient and adaptable network backbone for iterative refinement. Current options like U-Net and Vision Transformer often rely on resource-intensive deep networks and lack the flexibility needed for generating images at variable resolutions or with a smaller network than used in training. This study introduces LEGO bricks, which seamlessly integrate Local-feature Enrichment and Global-content Orchestration. These bricks can be stacked to create a test-time reconfigurable diffusion backbone, allowing selective skipping of bricks to reduce sampling costs and generate higher-resolution images than the training data. LEGO bricks enrich local regions with an MLP and transform them using a Transformer block while maintaining a consistent full-resolution image across all bricks. Experimental results demonstrate that LEGO bricks enhance training efficiency, expedite convergence, and facilitate variable-resolution image generation while maintaining strong generative performance. Moreover, LEGO significantly reduces sampling time compared to other methods, establishing it as a valuable enhancement for diffusion models.
more »
« less
Board 121: Development of a Create-a-Lego-Engineer Activity to Examine Students’ Engineering Identity
This paper describes the development of a Create-a-Lego-Engineer (CALE) activity which was created as an alternative to the Draw-a-Scientist (DAST) and Draw-an-Engineer Tests (DAET). While the DAST and DAET examine students’ (mis)conceptions of scientists and engineers, they provide limited information about whether students can envision themselves as scientists or engineers now or in the future. We drew from the Lego Serious Play (LSP) method which is grounded on the premise that hands-on learning results in a deeper understanding of the world and oneself in it. The LSP method is a process used to enhance innovation and business development, and it involves adults building metaphorical representations of their identity using Lego bricks. We adapted this process for use with elementary students (3rd-5th grade) in a specific context, namely students are asked to build themselves as engineers and a scene depicting what they would be doing as an engineer. Lego bricks were chosen as they are familiar to most students, are easy to use even without prior experience, and require no special skills or artistic abilities. The activity allows us to explore students’ creations of physical representations of themselves as engineers, including issues related to gender and physical characteristics (e.g., skin color, hair color and style), all of which students can customize using a variety of Lego options. Students are provided with a variety of Lego person pieces in order to try and build a representation of themselves using Legos. Additionally, a wide variety of Lego brick pieces were provided in order to allow for numerous ways in which students might represent engineers doing engineering work. Students were asked to imagine themselves as engineers and then to create their Lego engineer. Next, on a notecard, they described the type of work their Lego engineer would be doing, at which point they were then asked to create this scene using Lego bricks. Finally, after completing their creations, students reflected on the meaning of what they built and verbally described their creation and the choices they made. While these reflections provide additional insight into students’ beliefs about who can be an engineer and what engineers do, they also provide students an opportunity to imagine and see themselves in the role of an engineer. This activity was developed within the context of a multi-year, NSF-funded research project examining the dynamics between undergraduate outreach providers and elementary students to understand the impact of the program on students’ engineering identity and career aspirations. This paper will describe the development of the activity as well as preliminary findings from pilot testing and use with elementary students participating in the overall research project. Potential implications and limitations will be described.
more »
« less
- Award ID(s):
- 1657509
- PAR ID:
- 10483952
- Publisher / Repository:
- ASEE Conferences
- Date Published:
- Format(s):
- Medium: X
- Location:
- Tampa, Florida
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
CONTEXT For the last 40 years, the aggregate number of women receiving bachelor’s degrees in engineering in the US has remained stuck at approximately 20%. Research into this “disappointing state of affairs” has established that “the [educational] institutions in which women sought inclusion are themselves gendered, raced and classed” (Borrego, 2011; Riley et al., 2015; Tonso, 2007). PURPOSE Our focus is women students who thrive in undergraduate engineering student project teams. We need to learn more about how they describe becoming an engineer, about how women come to think of themselves as engineers and about how they perform their engineering selves, and how others come to identify them as engineers (Tonso, 2006). METHODS We are guided by a feminist, activist, and interpretive lens. Our multi-case study method, i.e., three semi-structured interviews and photovoice, offers two advantages: 1) the knowledge generated by case studies is concrete and context dependent (Case and Light, 2011); 2) case studies are useful in the heuristic identification of new variables and potential hypotheses (George and Bennett, 2005). ACTUAL OUTCOMES Our preliminary results suggest these women find joy in their experience of developing and applying engineering expertise to real, tangible, and challenging problems. They find knowing-about and knowing-how exciting, self-rewarding and self-defining. Further, these women work to transform the culture or ways of participating in project teams. This transforming not only facilitates knowing-about and knowing-how; but also it creates an environment in which women can claim their expertise, their identity as engineers, and have those expertise and identities affirmed by others. CONCLUSIONS If we aim to transform our gendered, raced, classed institutions, we need to learn more about women who thrive within those institutions. We need to learn more about the joy of doing engineering that these women experience. We also need to learn more about how they create an “integration-and-learning perspective” for themselves (Ely and Thomas, 2001) and a “climate for inclusion” within those project teams (Nishii, 2012), a perspective and climate that fosters the joy of doing engineering.more » « less
-
Engineers are responsible to many stakeholders, including the public and their employer. One such responsibility is considering and accounting for the potential impacts and risks associated with a technology that they create. A relatively new, and potentially risky, technology that has been on the rise over the past two decades is social media. The advent of social media, such as Facebook and Twitter, and its integration into our daily lives raises questions about the duty engineers bear for its responsible usage and design versus the responsibilities users have as they use the technology. This paper analyzes qualitative interview data from a study on engineering students’ perceptions of engineering ethics and social responsibility to answer the following research question: In what ways do students change (or not change) how they talk about engineers’ social and professional responsibilities to the technologies they create when framed in the context of social media? Our findings show that mentioning social media as a specific application of engineering ethics rendered visible the relationship between engineers, users, and technology that students then utilized to address the broader question about engineers’ responsibility to the technologies they create. In this study, a total of 33 students from three U.S. universities were interviewed longitudinally, once in the first year of their degree and again in the fourth year. In the interviews, the students were asked about their views on the social and professional duties engineers have for the technologies they create, framed in the context of social media. Analysis of student responses involved open and axial coding of relevant interview portions performed by two researchers to identify common themes and longitudinal changes between student interviews. These themes included: communication between the engineer and user, collective responsibility, benefits to society, high quality engineering, and misinformation. While students typically maintained elements of their views across both interviews, it was also common to see students change their responses to include new themes or exclude themes present in their initial interview. The students tended to believe that engineers have a responsibility to think through potential uses (or misuses) of their technology, but also believe that the users share some responsibility to use the technology appropriately. When social media was mentioned specifically, some students believed that the users were entirely responsible for how the technology is used, occasionally contradicting their views of engineering ethics when probed without the context of social media. This paper highlights the central tension between user responsibility and engineer responsibility. By illuminating students’ views, it will support educators in opening a dialogue with their students about who is ultimately responsible for the design and use of new technologies.more » « less
-
The purpose of this complete research paper is to analyze the impacts of an open makerspace on the development of students’ engineering identities. This paper seeks to build upon current belonging analyses about makerspaces and shift the focus towards students’ engineering identities. Our team interviewed 17 first-year engineering students attending a small, private university located in the American southwest. During the interviews, they were asked to reflect on their experiences in classes and involvement in engineering related activities. Some of the interview questions are influenced by previous models of engineering identity. Our research team noticed a pattern of students spending personal time using the Makerspace in their engineering department. This is an open workshop where students have access to free supplies to do what we’ve called “make” which is the act of problem solving, designing, and building using the tools provided. The high rate at which this space is mentioned in tandem with the students’ successes during the two semesters exemplifies the impact it has on student retention rates. We noticed a trend that students who have strong engineering identities tend to spend time making in the Makerspace. Any mention of the Makerspace itself or any connective context pieces relating to activities of the Makerspace spoken by the group of students were collected by our research team. This paper will examine how heavy of an impact, if at all, the Makerspace has on the further development of a student's ability to recognize themselves as an engineer if they came into college with an initial interest in making. Our analysis suggests the Makerspace provides an opportunity for students to display performance when making. This in turn causes students to see themselves as engineers when they experience internal and external recognition from being in the Makerspace. The results of this analysis will aid in the creation of effective intervention methods universities can implement during the first year engineering curriculum to increase retention rates.more » « less
-
nterest in science, technology, engineering, and mathematics (STEM) begins as early as elementary and middle school. As youth enter adolescence, they begin to shape their personal identities and start making decisions about who they are and could be in the future. Students form their career aspirations and interests related to STEM in elementary school, long before they choose STEM coursework in high school or college. Much of the literature examines either science or STEM identity and career aspirations without separating out individual sub-disciplines. Therefore, the purpose of this paper is to describe the development of a survey instrument to specifically measure engineering identity and career aspirations in adolescents and preadolescents. When possible, we utilized existing measures of STEM identity and career aspirations, adapting them when necessary to the elementary school level and to fit the engineering context. The instrument was developed within the context of a multi-year, NSF-funded research project examining the dynamics between undergraduate outreach providers and elementary students to understand the impact of the program on students’ engineering identity and career aspirations. Three phases of survey development were conducted that involved 492 elementary students from diverse communities in the United States. Three sets of items were developed and/or adapted throughout the four phases. The first set of items assessed Engineering Identity. Recent research suggests that identity consists of three components: recognition, interest, and performance/competence. Items assessing each of these constructs were included in the survey. The second and third sets of items reflected Career Interests and Aspirations. Because elementary and middle school students often have a limited or nascent awareness of what engineers do or misconceptions about what a job in science or engineering entails, it is problematic to measure their engineering identity or career aspirations by directly asking them whether they want to be a scientist/engineer or by using a checklist of broad career categories. Therefore, similar to other researchers, the second set of items assessed the types of activities that students are interested in doing as part of a future career, including both non-STEM and STEM (general and engineering-specific) activities. These items were created by the research team or adapted from activity lists used in existing research. The third set of items drew from career counseling measures relying on Holland’s Career Codes. We adapted the format of these instruments by asking students to choose the activity they liked the most from a list of six activities that reflected each of the codes rather than responding to their interest about each activity. Preliminary findings for each set of items will be discussed. Results from the survey contribute to our understanding of engineering identities and career aspirations in preadolescent and adolescent youth. However, our instrument has the potential for broader application in non-engineering STEM environments (e.g., computer science) with minor wording changes to reflect the relevant science subject area. More research is needed in determining its usefulness in this capacity.more » « less
An official website of the United States government

