Avatar

Ekaterina (Katja) Sirazitdinova

Research Engineer in Computer Vision and Machine Learning, Data Scientist @ NVIDIA

Hi! I am Ekaterina (or Katja)

I am a research engineer focusing on computer vision and deep learning. During my PhD studies, I was mainly working on the tasks of image-based 3D reconstruction and tracking, and afterwards spent some time doing research in deep learning for medical image analysis. I work as a data scientist at NVIDIA now, helping customers to solve their computer vision and video analytics problems. I also specialize in machine learning inference optimization and deployment of AI models on data center GPUs as well as in embedded / edge devices.

Interests

  • Deep Learning
  • Computer Vision
  • 3D Reconstruction
  • Inference Optimization

Education

  • PhD in Medical Image Processing, 2018

    RWTH Aachen University, Germany

  • MSc in Media Informatics, 2014

    RWTH Aachen University, Germany

  • Diploma in Business Informatics, 2011

    Baikal State University of Economics and Law, Russia

Experience

 
 
 
 
 

Data Scientist for Computer Vision, Video Analytics and Deep Learning

NVIDIA

Mar 2020 – Present Munich, Germany
I help customers to solve their computer vision and video analytics problems. I also specialize in machine learning inference optimization and deployment of AI models on data center GPUs as well as in embedded / edge devices.
 
 
 
 
 

Research Engineer Machine Learning and Data Science

Konica Minolta Laboratory Europe

May 2018 – Feb 2020 Munich, Germany

Utilising deep learning in tasks of medical image processing:

  • Nuclei segmentation in H&E stained digital pathology images.
  • Lung segmentation in X-ray images.
  • Detection of abnormalities in orthopaedic X-ray images.
  • Classification of X-ray images according to body regions.
  • Segmentation of cancerous nodules in 3D (CT scans).

Python, PyTorch

Product conceptualisation in the area of computational precision medicine and implementation of an interactive prototype.

JavaScript, Quasar, Node.js, Dash, Python

 
 
 
 
 

Research Intern

National Institute of Informatics (Computer Vision Group)

Sep 2015 – Dec 2015 Tokyo, Japan

Development and implementation of an algorithm for probabilistic 3D fusion of depth maps acquired with a Microsoft Kinect sensor to enable an accurate 3D reconstruction with a better preservation of fine features compared to state-of-the-art approaches. Results are published in the 3DV conference proceedings.

C++, OpenCV, PCL, OpenGL, OpenCL, CUDA

 
 
 
 
 

Senior Researcher (Computer Vision and Image Processing)

Uniklinik RWTH Aachen (Department of Medical Informatics)

Oct 2014 – Mar 2018 Aachen, Germany

Researcher and developer at the projects:

  • IMAGO: image-guided system for outdoor navigation of visually impaired people.
  • Optical measurement of water overflows in urban drainage structures; an image-based system for automatic measurement and tracking of volumetric flow in sewers.
  • Ocular redness quantification: a tool for quantitative assessment of eye redness in clinical trials.
  • Documentation of chronic wounds in 3D.
  • Boundary aware teeth segmentation in dental fluoroscopy images.

C++, OpenCV, Java, MATLAB, Objective-C

Lecturer in the courses:

  • Methodology in Medicine for students with technical background (SS 2017, WS 20172018)
  • Medical Image Processing (SS 2017)
  • Image Processing and Handling (WS 20172018)

Tutor in the exercises:

  • Medical Image Processing (SS 2016, SS 2017)
  • Image Processing and Handling (WS 20162017, WS 20172018)
  • Supervision of bachelor and master theses.
 
 
 
 
 

Software Developer

Pilas-IT

Feb 2014 – Jun 2014 Aachen, Germany

Development of a web-based mobile community network application.

JavaScript, Parse, Facebook API, CSS3, HTML5, Bootstrap

 
 
 
 
 

Student Research Assistant

Uniklinik RWTH Aachen (Department of Medical Informatics)

Oct 2013 – Mar 2013 Aachen, Germany

iOS applications development:

  • Application for image acquisition in clinical trials (a tool to compare eye redness before and after putting in eye drops).
  • A software enabling data exchange between MATLAB server and iOS client app.

Objective-C, MATLAB

 
 
 
 
 

Research Intern

National Institute of Informatics (Interaction Analysis)

Mar 2013 – Sep 2013 Tokyo, Japan
An eye-tracker based user study focused on eye movement patterns in conversational process.
 
 
 
 
 

IT Assistant

TransTeleKom Company (TTK-Baikal)

May 2007 – Aug 2011 Irkutsk, Russia

Designing of an automated system for customer relationship oriented business processes (Diploma thesis).

Automation of business processes.

Database maintenance.

Visual Basic, 1C

MOOC Certificates

Latest Publications

(2019). Validation of Computerized Quantification of Ocular Redness. Translational Vision Science & Technology, 8(6).

DOI

(2019). A Multi-organ Nucleus Segmentation Challenge. IEEE Transactions on Medical Imaging.

Dataset Project DOI

(2019). Revisiting Depth Image Fusion with Variational Message Passing. In 3DV.

Code Dataset DOI

(2019). Removal of Simulated Biofilm at Different Implant Crown Designs with Interproximal Oral Hygiene Aids: An In Vitro Study. Clinical Oral Implants Research, 30(7).

DOI

(2018). Sewer Discharge Estimation by Stereoscopic Imaging and Synchronized Frame Processing. Computer‐Aided Civil and Infrastructure Engineering, 33(7).

DOI