×
Community Blog Alibaba Cloud MVP Story - Then Arifin

Alibaba Cloud MVP Story - Then Arifin

In this blog, MVP Then Arifin shares his experience of developing face recognition system and his perspectives on the role of AI and utilizing cloud native with Alibaba Cloud.

As an Alibaba Cloud MVP, Tech blogger, Certified Cloud Professional, Software Solution Architect and Engineer, Then Arifin's current role is an IT Analyst from one the Top Global Fortune companies. His current role focuses on application development and modernization via cloud computing adoption in Singapore. He would like to share his story about Artificial Intelligence (AI) and his interest in Cloud Native world with us.

Face Recognition System

During my college years, my main interest was to design face recognition system as our final thesis together with my two friends and the title of our thesis is "Analysis and Design Face Recognition System Using the Method of Discrete Cosine Transform (DCT) Energy Histogram".

The DCT itself is the method which is used for image compression like JPEG image format. The evaluation made use of the real-time images and non-real-time images. As a starting point, we used the real-time images by developing a system to capture individual images using webcam as a query image. Thereafter, we program the system to do the comparisons and evaluations from the face databases which we captured and recorded extensively from 20 individuals by collecting 10 to 12 face mimics per individuals. For further comparison, we used non-real-time images. We worked using Yale Face database as a query image and database image to run testing and simulation.

All facial images are taken in a frontal position with grayscale quality image. The recognition process that we used was Forward 2D-DCT calculation in order to obtain the DCT coefficient which was useful for valid feature vector selection. To recognize a face image, the system compared the image's valid feature vector from query image to each of the valid feature vectors in the database using the Euclidian distance classifier. As a result of different layers of comparisons and extensive evaluations, our system not only able to recognize 88.64 % for real time images but the system itself also enabled to recognize 61.82 % for non-real time images successfully.

1
Figure 1. Face Recognition Process Flow Using DCT Energy Histogram

Journey to The Cloud

Artificial Intelligence (AI) and Cloud Computing has increasingly becoming a great demand as an essential top skill in today's environment. My Cloud journey started in 2018 by taking self-paced online training and cloud-skill certifications. Eager to learn more about Alibaba Cloud services, I signed up as a cloud user on Alibaba Cloud console and academy in 2019. Through Alibaba Cloud Academy, I enrolled to "Cloud and Proud" program and successfully achieved Cloud and Proud Hero with the highest score qualification in July 2019.

My objective for this year is to complete my learning paths at Alibaba Cloud Academy, in particularly Machine Learning (ML) Engineer learning path and also focusing on exam preparations for certification in ACA and ACP Cloud Computing this year. According to Statista, the global value of the AI market will surpass more than an estimated $126 billion annually by 2025. A significant percentage of that value will occur as artificial intelligence powers cloud computing to help streamline workloads and automate repetitive tasks, improving data analytic and management, and integrate to Software-as-a-Service (SaaS) services to deliver more value.

AI and cloud computing are transforming business at every level. From deeper learning to near-compete automation of key processes, the potential is promising. In this present-day, Artificial Intelligence (AI) has collaborated with Cloud Computing, for instance, the AI features like text analytics, speech recognition, facial recognition, image recognition, language translation, intelligence assistant are accessible to the developers. They are be able to integrate these AI features into their development projects and research. These AI features in fact make it quite straightforward compare to my thesis in my college time when we need to design a facial recognition system in a complex manner.

By using Alibaba Cloud platform, it is possible for everyone to use and enjoy these AI features via Alibaba Cloud Intelligence Brain service. One of the most interesting scenarios from this Intelligence Brain service is the implementation of Aviation Brain at Hangzhou Xiaoshan International Airport for facial image recognition at immigration check-point. The process of recognition it claims only takes 0.3 seconds which is really fast and accurate.

Proud as Alibaba Cloud MVP

It is a challenging year for all of us in 2020 and everyone is hoping for the Covid-19 pandemic to be over soon. I have been nominated as an Alibaba Cloud MVP this year and this has become part of the achievements to my Cloud journey. As Alibaba Cloud MVP, it enables me to gain access and input to and from the Product Groups at Alibaba Cloud via technical sessions with Alibaba Cloud experts, Alibaba Cloud MVP community and special trainings about the services.

Furthermore, becoming Alibaba Cloud MVP will enable me to find out about the latest update of the product suites and services. It also assist me to fully understand about the best practices of using and managing the services and find out their focuses and the background of Alibaba Cloud particular approaches. These insightful knowledge will continue to support me to guide the community via social technology forums and blogs. Like many other Alibaba Cloud MVPs with their success stories, the knowledge that I have gained will enable me to build cloud native applications using Alibaba Cloud services for innovation and better solutions.

0 0 0
Share on

Alibaba Cloud MVP

40 posts | 33 followers

You may also like

Comments

Alibaba Cloud MVP

40 posts | 33 followers

Related Products