CV

This is a description of the page. You can modify it in '_pages/cv.md'. You can also change or remove the top pdf download button.

Contact Information

Name Joshua Peter Ebenezer
Professional Title Staff Research Engineer
Email joshuapeterebenezer@gmail.com

Professional Summary

Staff Research Engineer with 6+ years building production-grade deep learning systems for computational photography and video quality. Led AI models deployed on millions of devices; 5 US patents, CVPR/TIP/ICIP/ICASSP publications, and direct ownership of core camera pipelines.

Experience

  • 2025 -

    San Francisco, CA

    Staff Research Engineer
    Samsung Research America
    Leading end-to-end development of production AI models for flagship Samsung camera systems deployed across 30+ million devices.
    • Led development of Telephoto (Tele1), Max Mode, and Extreme Low-Light Night Mode pipelines
    • 2 US patents filed for Night Mode innovations
    • CVPR 2025 publication on advanced camera algorithms
    • Model architecture, training strategy, and data processing ownership
  • 2023 - 2025

    San Francisco, CA

    Senior Research Engineer
    Samsung Research America
    Developed AI systems for Super-Resolution and Night Mode imaging in telephoto cameras.
    • Owned image registration and alignment systems for Super-Resolution software
    • Developed novel ground truth rendering pipelines and specialized loss designs
    • 2 US patents filed for imaging algorithms
  • 2020 - 2021

    Seattle, WA

    Applied Science Intern
    Amazon Prime Video
    Designed large-scale video quality studies and developed deep learning methods for real-time video analysis.
    • Conducted subjective video quality study with 750 videos and 95 participants
    • Developed deep learning methods to detect audio-video synchronization errors in live sports
    • Informed bitrate ladder design for streaming services

Education

  • 2019 - 2023

    Austin, TX

    MS + PhD
    University of Texas at Austin
    Electrical & Computer Engineering
    • Advisor: Prof. Alan Bovik
    • Thesis: No-Reference Video Quality Assessment for Livestreamed HDR and High-Motion Videos
    • Sponsored by Amazon Prime Video
  • 2015 - 2019

    Kharagpur, India

    B.Tech
    Indian Institute of Technology (IIT) Kharagpur
    Electronics & Electrical Communication Engineering
    • Minor in Computer Science and Engineering
    • Institute Rank: 3 out of 643 students
    • GPA: 9.77/10
    • Nilanjan Ganguly Memorial Award for Best Undergraduate Thesis

Skills

Deep Learning: PyTorch, TensorFlow, Diffusion Models, GANs, Multi-frame Super-Resolution
Computational Photography: Image Registration, HDR Imaging, Night Mode Development, Low-light Imaging, Super-Resolution
Image & Video Quality: No-Reference VQA Metrics, Subjective Quality Studies, Streaming Video Evaluation, HDR Quality Assessment
Data & Training: Large-scale Data Processing, Loss Function Design, Evaluation Pipelines, Model Architecture Design