AI

Challenges: Knowledge is humanity’s greatest asset. Books are invented to keep the knowledge is well-shared from generation to generation. However, in the digital transformation age, we need to transfer the data from books to digital databases. If you are looking for a method to transform the information from papers, books, or even you want to build a solution to help people to copy the information on what they see by using the smartphone application, Smart OCR is the solution that BAP can provide for those business purposes Solution: SmartOCR is an application that provides the functionality to read the text from any pictures. End-user can simply drop the pictures to the application, it will automatically read the text from pictures and copy it, end-user can paste that text to wherever they want. Moreover, end-user can use the smartphone’s camera to zoom to what they want to read, SmartOCR will do its job perfectly by reading the text and copy it to end-users, the function is quite simple and easy to use but there is a smart system behind to make it work perfectly. We are proud of the systems that we have researched and developed because there are a lot of new technologies such as AI, Big data, GPS, Maps….are implemented. On the other hand, we also listen to get more requirements from customers to build a suitable system for any business needs. Send us a message to sales@bap.jp then we can discuss more and explore new adventure together

Auto Scoring Input of this App is an answer sheet which have tic & toe in the answer boxes. This will be processed with Image Processing and Objects Detection with Yolov5. Beside of that, it also Detect Student Code and Examination Code on the answer sheet to get Students Information. Problem 1: Data We must have to create data by ourself, because the type of this data not have in any Public Dataset on the Internet. Problem2: Input images size Because we need to capture full A4 paper, so the images size are very important to me. If the input images size are very low, so we can’t extract informations from the answer sheet. And it will affect to the accuracy of my App. Problem2: Results

Problem: Image Captioning Given an image, our goal is to generate a caption. Input: Image Output: Caption for image Solution For this problem, we use will use InceptionV3 (which is pre-trained on Imagenet) to classify each image. We will extract features from the last convolutional layer. The RNN (here GRU) attends over the image to predict the next word. Experimental Results

Challenges: The modern people nowadays are always busy, they don’t even have time to go shopping for their necessities. On the other hand, fashion retailers are facing abandoning baskets when it comes to online shopping and huge amounts of returned items to the stores because of mis-sized or color issues. SMART FASHION is born to help fashion retailers solve that problem by applying AI technology into trying clothes on virtually and suggesting matching items with the user’s current clothing in their wardrobe. Solution: SMART FASHION is a new weapon to get rid of this distance. On this app, users are able to use the virtual fitting room to try on clothes, furthermore, they can style themselves from their own wardrobe with suggesting items on the app, Smart Fashion will then score the matching outfits created by end users . This is a solution for time-saving and enhancing user experience on E-commerce, or retailers. In order to do that, machine learning plays the key role in developing this app, when people can feel like they are real on each clothing items. This can be considered as one of the best developments by BAP’s team. BAP is a company that provides solutions for businesses . Besides the written down requirements from customers, we are ready to give ideas to contribute to the best performance. Our engineers have a lot of experience in Japanese’s companies and have worked with Japanese clients, so we can provide ideas which fit with Japanese culture as well. Besides AI technology, we can also provide other technological services – Blockchain, Web Application, Big Data, and Game, or and ERP… Development Form Lab Development Language AI Deep Learning Machine Learning Product Web, IOS, Android Development Site BAP Software

Overview Sharingan is a Smart System with core technology is facial identification by Artificial Intelligence. The system provides 2 main services: Automatic Attendance Automatic Surveillance Demo Automatic Attendance Fast Accurate identification Easily deploy and leverage enterprise infrastructure Can time attendance many people at the same time Put an end to fake and forged timekeeping Automatic Surveillance Supervise employees in the workplace of the enterprise Supervise students in the classroom and notify parents

Problem: The data on social platform are growing everyday. How can we use such data to discover “What is happening” on Facebook or Twitter everyday or even every hour? We can use the topic modeling for this problem: – INPUT: A dataset of Facebook’s posts of the most popular fanpages in Vietnam at a specific time. – OUTPUT: A set of top hottest topics that are the most popular on Vietnamese Facebook at that time and their keywords. Solution: For this problem, we use the LDA Topic Modeling Technique to solve the challenge. LDA is a probability model that find the pattern topic distribution in a corpus (a set of documents). Result: In October 2019, we discovered these are the hottest topics that Vietnamese Facebook-er were talking about the most: We also can present the fanpages by vectorizing them, and compare them to look for the similarities: Here is an comparison between pages base on their content in Oct-2019. The more blue a square is, the more similar two fanpages are: Oppoturnity: With a proper crawler, an improved Topic Modeling technique, we can track the hot topics on Facebook or any social network in every hour or even every second.