AI/ML Engineer

URINOV AZIZBEK

한국인식산업(주) (2023 - 2025), Guro, Seoul - South Korea

About me

My name is Urinov Azizbek, and I’m from Uzbekistan. I have a deep passion for technology and innovation, which led me to pursue my studies in Information Technology in South Korea.

I'm currently working as an AI/ML Engineer with over two years of experience in building intelligent systems. My main focus is on researching and experimenting with state-of-the-art AI models, always staying up to date with the latest advancements in the field. I enjoy solving complex problems with data-driven approaches and continuously learning to push the boundaries of what's possible with AI.

경기, 안양시 - South Korea
Chartak, Namangan, Uzbekistan
남 - 27세 - 1998년
010-5730-7797
uacoding01@gmail.com
https://github.com/zero-suger
한국인식산업(주) (2023 ~, AI/ML Engineer)

iBETA Level 1 [Liveness Detection]

Passed iBETA Level 1 Exam on 2025.05.07 (APCER 0%) using passive detection on RGB images only, 2nd in South Korea.

verified Certificate Link

8 SOTA models [Face Detection]

Face Detection with DLib, Haar-Cascase, Mediapipe, MTCNN, RetinaFace, SCRFD, SSD, YOLO, YuNet, etc.

link GitHub Link

Liveness Detection TEST [Vercel]

A Liveness Detecting web application by using Face Detection + Face Liveness Detection (Lightweight models).

link TEST LINK [Click me]

MiniCPM-V Django Web App [LLM]

A web application leveraging MiniCPM-V, a multimodal large language model for image and text processing by using DJANGO.

link GitHub Link

Manipulation Localization [FMDL]

Detecting OpenCV, Photoshop, Diffusion, StyleGAN, and Online Tools Manipulated Face images and Localize that area.

Warning Restricted to share

Face Morphing/Recognition [FMR]

Morphing two IDs using Diffusion, GAN, and OpenCV methods, and evaluating the impact on face recognition.

Warning Restricted to share

Learning AI in 2025 [my_roadmap]

Starting with Python and Python libraries, then progressing to Machine Learning, DL (with a focus on Computer Vision), and LLMs.

link GitHub Link

Generating/Detecting DeepFakes

Deepfakes were generated using DeepFaceLab by altering facial identities in videos, followed by analysis of detection techniques like artifact inspection and deep learning classification.

youtube_activity Watching video