Academy of Interdisciplinary Studies Division of Integrative Systems and Design 214 Generative AI in 3D Food Printing Supervisor: LI Mitch / ISD Student: KANANI Muhammad Zohair / COMP Course: UROP 1000, Summer The desire to know what we are eating is increasing every day. Even if we exclude athletes and aged people who have been strictly advised to a particular food guide, a normal person who visits gym one hour a day is also conscious of his nutritional intake. While packaged food have a detailed table showing exact figures for each nutrition, the food we cook or eat at restaurants lack these numbers. Through our research, we investigate the importance of deep learning in classifying food and extend it to segment food images to estimate the values that can help in calorie-controlled 3d food printing and building personal food recommendation system. Multi-axis DLP 3D-Printing for Applications in Soft Robotics Supervisor: SCHARFF Rob Bernardus Nicolaas / ISD Student: WAN Ho Cheung / QSA Course: UROP 1000, Summer This project is named Multi-axis DLP 3D-Printing for Applications in Soft Robotics. This project and report aimed to explore the potential exploitation of multi-axis DLP (Digital Light Processing) 3D-printing in the field of Applied Soft Robotic. This report will consist of an overview of SLA (Stereolithography), DLP, Grayscale DLP (g-DLP), multi-axis DLP 3D-printing, Soft Robotic and GRACE (GeometRy-based Actuators that Contract and Elongate). Then, a discussion about the potentialities of the multi-axis DLP 3D-printing method’s application on the field of soft robotics that the traditional 2.5D printing method cannot offer. Such potentialities include printing GRACE in a support-free manner and printing a curved GRACE with the utilization of grayscale material. LLM for Networking Supervisor: SONG Shenghui / ISD Student: SHEN Yuming / ISD Course: UROP 1000, Summer This report focuses on the application of Large Language Models (LLM) for Networking resource allocation. LLM for Networking is gradually becoming a new research direction in today's booming development of generative AI. There are also not many articles on this direction but the growth rate is extremely fast. At this stage, it is essential to propose a variety of different implementation ideas and to demonstrate their advantages over traditional network resource allocation through comparison. We will summarise and compare existing works so that we can sort out our future research directions and make relevant research plans.
RkJQdWJsaXNoZXIy NDk5Njg=