Abhishek Tiwari, Developer in Bengaluru, Karnataka, India
Abhishek is available for hire
Hire Abhishek

Abhishek Tiwari

Verified Expert  in Engineering

Image Processing Developer

Location
Bengaluru, Karnataka, India
Toptal Member Since
July 12, 2021

Abhishek is a seasoned computer vision and AI engineer with nearly a decade of experience in enterprise organizations, including GE, Samsung, and Philips as well as startups. He's skilled in a number of languages (Python, C++), operating systems (Linux, Windows), libraries (PyTorch, TensorFlow, Keras), and techniques (feature extraction, segmentation). Abhishek's past work includes camera images (2D + time, 2D + depth + time), ultrasound images (2D, 3D, 3D + time), and MR images (3D).

Computer VisionMachine LearningDeep LearningArtificial Intelligence (AI)PythonOpenCVNumPyC++PyCharmScikit-learnEdge ComputingLinuxDockerMATLABAmazon Web Services (AWS)Caffe

Portfolio

Vimaan Robotics
Python, PyTorch, OpenCV, NVIDIA TensorRT, Intel RealSense, Docker, Linux...
GE Healthcare
3D Image Processing, C++, Computer Vision, Docker, Python, Linux...
Samsung
3D Image Processing, C++, Computer Vision, Image Processing, ITK, Python, NumPy

Experience

Computer Vision - 7 yearsImage Processing - 7 yearsOpenCV - 5 yearsPython - 5 yearsC++ - 4 yearsITK - 4 yearsDocker - 2 yearsPyTorch - 1 year

Availability

Part-time

Preferred Environment

Python, C++, Image Processing, Computer Vision, OpenCV, ITK, PyTorch, Keras, Intel RealSense, Docker

The most amazing...

...thing I've worked on was an entire stack including image analysis algorithms, machine and deep learning, visualization, and basic rendering in a single project.

Work Experience

Module Lead - Computer Vision

2020 - PRESENT
Vimaan Robotics
  • Developed algorithms for object detection and instance segmentation and semantic segmentation problems for images using CNN.
  • Implemented fusion algorithms across different data modalities such as RGB and depth data.
  • Optimized deep learning models using TensorRT and deployed on Edge Devices.
  • Containerized and deployed algorithms using Docker.
  • Provided technical guidance to team members in algorithms such as image classification and text detection and recognition (OCR).
Technologies: Python, PyTorch, OpenCV, NVIDIA TensorRT, Intel RealSense, Docker, Linux, Image Processing, Computer Vision, Edge Computing, Google Cloud Platform (GCP), Machine Learning, Deep Learning, Convolutional Neural Networks (CNN), Amazon Web Services (AWS)

Senior Software Engineer

2018 - 2020
GE Healthcare
  • Contributed towards development and commercialization of AIRx Brain application targeted towards intelligent magnetic resonance imaging (MRI) scanner.
  • Mentored an intern and new-joinee towards the application of artificial intelligence technologies towards other MRI applications.
  • Solved multiple critical issues with high impact on GE MR business.
Technologies: 3D Image Processing, C++, Computer Vision, Docker, Python, Linux, Image Processing, Machine Learning, Deep Learning, Convolutional Neural Networks (CNN)

Technical Lead

2018 - 2018
Samsung
  • Researched and developed the image segmentation and quantification algorithms for 2D musculoskeletal ultrasound datasets.
  • Developed bone structure matching algorithm for 2D musculoskeletal ultrasound datasets.
  • Granted a world patent for being a co-contributor to algorithm applications to medical imaging.
Technologies: 3D Image Processing, C++, Computer Vision, Image Processing, ITK, Python, NumPy

Lead Engineer

2016 - 2018
Samsung
  • Automated user inputs for image segmentation algorithm using artificial intelligence algorithms.
  • Received a merit award for presentation of team's work at Samsung Best Paper Award (SBPA) internal conference among all Samsung's worldwide R&D centers.
  • Presented the algorithm at SPIE Medical Imaging Conference titled "HoDOr: histogram of differential orientations for rigid landmark tracking in medical images."
Technologies: 3D Image Processing, C++, Caffe, Computer Vision, Image Processing, ITK, Scikit-learn, Machine Learning, Deep Learning, Convolutional Neural Networks (CNN)

Senior Software Engineer

2014 - 2016
Samsung
  • Researched and developed image segmentation and quantification algorithms for 4D cardiac ultrasound datasets (3D + time).
  • Implemented visualization algorithms for succinct demonstration of algorithm output and quantification metrics.
  • Presented the algorithm at the IEEE EMBC 2016 Conference titled "Mitral Valve Annulus Localization in 3D Echocardiography."
Technologies: Python, C++, 3D Image Processing, Computer Vision, ITK, VTK, Scikit-learn, Image Processing, MATLAB

Summer Research Intern

2012 - 2012
Philips
  • Developed an algorithm to delineate lesion boundary from ultrasound (b-mode) and elastography (strain) datasets.
  • Extracted intensity and strain-based features for lesion classification into benign or malignant.
  • Validated algorithm output against manual ground-truth annotation by a radiologist.
Technologies: Image Processing, MATLAB, OpenCV

Masters Thesis - Video Stabilization

https://github.com/tiwaria9034/tiwaria9034.github.io/blob/master/Files/masters_thesis.pdf
Developed a real-time digital video stabilization algorithm. The algorithm is based on two-dimensional feature-based motion estimation. The method tracks a small set of features and estimates the movement of the camera between consecutive frames. An affine camera motion model is used to determine the parameters of translation and rotation between images. The trajectory of computed affine parameters is temporally smoothed by using a moving average filter to remove high-frequency jitter, and motion compensation is performed based on smoothed trajectory resulting in improved quality of the video. The proposed algorithm stabilizes horizontal and vertical panning and rotations and is suitable for various applications such as handheld camcorder video, camera mounted on car video, cellphone camera video, etc.

MOOC Courses

https://github.com/tiwaria9034/Massive-Open-Online-Courses
The Github URL captures the code for the assignments as part of the MOOC courses attended. The courses are from learning sites such as Coursera, Udemy, Udacity, etc. The enrolled courses are:
1. [Udemy] Deep Learning: GANs and Variational Autoencoders
2009 - 2014

Bachelor's Degree in Electrical Engineering

Indian Institute of Technology, Kanpur - Kanpur, Uttar Pradesh, India

2009 - 2014

Master's Degree in Electrical Engineering

Indian Institute of Technology, Kanpur - Kanpur, Uttar Pradesh, India

Libraries/APIs

OpenCV, NumPy, Scikit-learn, PyTorch, Keras, VTK, TensorFlow

Tools

ITK, PyCharm, MATLAB

Languages

Python, C++

Frameworks

Caffe

Platforms

Docker, Linux, Google Cloud Platform (GCP), Amazon Web Services (AWS)

Other

Image Processing, Computer Vision, 3D Image Processing, Machine Learning, Deep Learning, Convolutional Neural Networks (CNN), Artificial Intelligence (AI), Computer Vision Algorithms, Intel RealSense, NVIDIA TensorRT, Edge Computing

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring