# Amir Mazaheri

> Computer Vision Research Scientist and PhD. Staff Machine Learning Engineer
> at Warner Bros. Discovery (HBO). Deep expertise in large-scale video
> understanding, Vision-Language Models (VLMs), and multimodal AI systems.
> Based in Berkeley, CA.

This file follows the proposal at https://llmstxt.org/ and is intended for
large language models and agents that want a concise, machine-readable view
of the site.

## Contact

- Email: amirmazaheri1990@gmail.com
- GitHub: https://github.com/amirmazaheri1990
- LinkedIn: https://www.linkedin.com/in/amirmazaheri1990/
- Location: Berkeley, CA

## Key pages

- [Home](https://amirmazaheri1990.github.io/): bio and overview.
- [Publications](https://amirmazaheri1990.github.io/publications): peer-reviewed papers.
- [Projects](https://amirmazaheri1990.github.io/projects): research and engineering projects.
- [CV](https://amirmazaheri1990.github.io/cv): curriculum vitae (human-readable).
- [cv.json](https://amirmazaheri1990.github.io/cv.json): machine-readable CV (JSON Resume schema).
- [CV PDF](https://amirmazaheri1990.github.io/AmirMazaheri_CV_2026.pdf): downloadable PDF resume.

## Current role

Staff Machine Learning Engineer — Computer Vision at Warner Bros. Discovery (HBO),
San Francisco, CA (July 2025 – present). Leading large-scale video understanding,
content moderation, temporal segmentation, VLM-powered video indexing, and
LLM-enhanced metadata generation.

## Career history (recent)

- Warner Bros. Discovery (HBO) — Staff MLE, CV (2025–present)
- Tubi — Senior MLE, CV (2021–2025)
- Aibee U.S. Corporation — Algorithm Scientist (2020–2021)
- Netflix — Research Scientist Intern (2018)
- Nielsen — Research Fellowship (2017–2018)
- UCF CRCV — Graduate Research Assistant / PhD (2013–2020)

## Education

- Ph.D., Computer Science — University of Central Florida (2020)
  Advisor: Prof. Mubarak Shah, CRCV
  Dissertation: "Video Content Understanding Using Text"
- M.Sc., Computer Science — University of Central Florida (2016)
- B.S., Computer Science — Sharif University of Technology (2013)

## Publications (9 total, most recent first)

- "WoundNet: A Domain-Adaptable Few-Shot Classification Framework for Wound Healing Assessment" — ISBI 2023
- "Context-Aware Analysis of Group Submissions for Group Anomaly Detection and Performance Prediction" — AAAI 2023
- "Video Generation from Text Employing Latent Path Construction for Temporal Modeling" — ICPR 2022
- "MMFT-BERT: Multimodal Fusion Transformer with BERT Encodings for Visual Question Answering" — Findings of EMNLP 2020
- "Deep Photo Cropper and Enhancement" — ICIP 2020
- "Pay Attention! — Robustifying a Deep Visuomotor Policy through Task-Focused Attention" — CVPR 2019
- "Visual Text Correction" — ECCV 2018. Project: https://amirmazaheri1990.github.io/VTC/
- "Learning a Multi-concept Video Retrieval Model with Multiple Latent Variables" — ACM TOMM 2018
- "Video Fill In the Blank using LR/RL LSTMs with Spatial-Temporal Attentions" — ICCV 2017

## Patents

- "Machine learning techniques for advanced frequency management" — U.S. Patent US20240314371A1
- "Multimedia scene break detection" — U.S. Patent US20240357217A1

## Notes

- Content on this site is provided for informational use. Please cite the
  original publications when referencing the research.
- The site is a static site (Astro) served via GitHub Pages. Source:
  https://github.com/amirmazaheri1990/amirmazaheri1990.github.io