Skip to content
View ozxmn's full-sized avatar

Block or report ozxmn

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ozxmn/README.md

Mansur Ozaman

First-year computer science student at Nazarbayev University in Astana. I work on telling human-made content from machine-made content, mostly in Kazakh.

Research

Unchecked or Falsely Accused? Evaluating AI-Generated Text Detectors on Kazakh (2026, under review)

AINA-KK: 3,811 Kazakh passages written before ChatGPT and 5,597 machine passages from 9 generators. Five public English-trained detectors flagged up to 81.4% of the human passages. AINA-KK-Det, the detector released with the paper, flags 0.9% of human test passages.

paper · code · data · detector · reproduce on Kaggle · project page

Evaluating Deep Learning and Traditional Approaches Used in Source Camera Identification (SIDe'26, Springer STEAM-H)

Which of four smartphones took a photo? Hand-built JPEG/DCT features reached 94.7% accuracy, an ImageNet ResNet-18 reached 62.3%.

paper · code · data

Products

  • Authini: detection of AI-generated images and video
  • Choovy: digital menus for restaurants

Contact

ozaman.vercel.app · LinkedIn · Hugging Face · Kaggle · mansur.ozaman@nu.edu.kz

Pinned Loading

  1. aina-kk aina-kk Public

    AINA-KK: benchmark and detector for machine-generated Kazakh text

    Python 1

  2. edltauSCI edltauSCI Public

    Source camera identification on four smartphones: DCT/JPEG features vs PRNU vs CNNs. Code for the SIDe'26 paper.

    Python

  3. imgreco imgreco Public

    Desktop app that estimates which phone took a photo, combining JPEG, PRNU and CNN analysis. Research code: ozxmn/edltauSCI.

    Python

  4. voiceX voiceX Public

    Desktop app that recognizes students by voice, for classes and exams. MFCC features, SVM classifier, CustomTkinter GUI.

    Python