This project revolves around the development of a software application tailored for finance professionals. The software aims to streamline the conversion of a company’s 10-K financial report into a PowerPoint presentation using automation and AI. The presentation is enhanced with visually engaging charts, succinct summaries, and a chat feature for efficient information retrieval.
Many studies have been carried out to integrate multi-modal data into a global feature space. In such a dataset, heterogeneous data like text, images, and videos, could be accessed and processed in a uniform manner. However, the integration of multi-modal data also means the loss of information, which makes it necessary to find methods that can extract relevant information from the global dataset both effectively and efficiently. That is, the search results from the dataset should have good quality and can be obtained at a low time cost. In this project, we would like to compare both search quality and efficiency of several search methods in a dataset uniformly storing embedded caption-image pairs. Specifically, we used CLIP to pre-process the dataset into high-dimensional vectors. Then, we applied different search methods, such as Nearest Neighbors and various Faiss methods with different parameters, on text-to-image and image-to-image search. Finally, we utilized precision@k and NDCG as the metrics for measurement. The text or image to search for might not only be selected from the dataset but also could be arbitrarily generated. During our evaluation, we discovered the trade-off between search quality and efficiency. As a result, we found that the clustering Faiss built on inner product could reach the optimal balance.
We plan to construct an unmanned COVID-19 testing solution. Our main customers are organizations such as governments, schools, companies who want to provide affordable, easy and reliable testing services to associated people such as local citizens, students and employees. Upon agreement, we install our machines on customers’ sites. Our machine is capable of three testing methods: PCR Swab, PCR Saliva and Antigen Swab. Sample collection is completely automatic, and no people are required on site. Antigen samples can be analyzed directly inside the machine with results available in a short period of time. PCR samples will be collected by our employees and transported to our labs for further experiments. Test results will be reported to the customer organization, and we consider our process done. We also provide software interfaces, so that we can exchange data with the customer organization, such as user authentication info and test results.
The goal of this project is to design a 4-bit “Absolute-value Detector” with the minimum energy and worst-case delay of 800ps. Here “delay” refers to the worst-case propagation delay and “energy” refers to total energy drawn from VDD for a given input probability distribution. You may use gate sizing and supply voltage scaling as variables.
The project implements the classic tic-tac-toe game on the Spartan-6 FPGA. The game board size is 3×3, and two human players are involved. The board of the game will be displayed on a monitor via VGA. The player who has the turn can press 4 buttons, UP, DOWN, LEFT, or RIGHT to move a blinking cursor on the board. When the cursor is moved to the target grid, the player can press the ENTER button to record the move, and if the game is not ended yet, the other player gets the turn. During the game, the 7-segment display will show the current location of the cursor, and which player’s move is now pending. We also need to constantly check whether or not the game should be over. There can be three outcomes: tie, Player 0 wins, or Player 1 wins. When a game-over condition is met, the game ends automatically, and no new moves are recorded. The final outcome will be shown on the 7-segment display. The user can press the RESET button to start a new game with an empty board.
This paper will compare the economic factors behind choosing between an Intel x86 MacBook Pro and an Apple M1 Pro (ARM) MacBook Pro through the perspective of a full-time video editor, a user-base known for demanding performance from their PCs.
On 14 April 1912, RMS Titanic, one of the most well-known ships in that era and hitherto, struck an iceberg while cruising at full speed towards New York City. Failing to withstand the damage, the Titanic soon sank after midnight, killing more than 1,500 people on board. This report will introduce the general information about this accident and discuss some ethical issues behind it. It is intended for the general public who is interested in the incident but has no knowledge of shipbuilding, so only the management aspects of the failure will be included.
On Sept. 24, 2021, Chinese authorities officially announced a final and full-scale crackdown on cryptocurrency trading and mining businesses within its border. Almost immediately, one of the most popular crypto trading platforms in China, Huobi, announced that they would stop new registrations and gradually retire existing accounts by Dec. 31. The global crypto market fell […]