Anand is available for hire

Anand Ajmera

Verified Expert in Engineering

Computer Vision Developer

Vadodara, India

Toptal member since March 23, 2020

Expertise

Computer Vision Deep Learning Machine Learning NumPy Git Python OpenCV TensorFlow PyTorch Linux C++

Bio

Along with holding a master's degree in autonomous systems, Anand has worked within a wide range of machine learning (ML) domains, including self-driving cars, medical AI, and industrial AI automation. He gets a thrill when conducting ML research for new apps and seeing it through to production-ready code. Anand thrives in startup environments and has remotely helped startups build their products from scratch while handling an entire ML pipeline, from data curation to deployment.

Portfolio

Freelance (India)

TensorFlow, PyTorch, Python

Pathway AI (San Jose, USA)

Swift, OpenCV, C++, TensorFlow, Python

Fraunhofer Research Institute (Germany)

C++, Python, TensorFlow

Experience

Computer Vision - 5 years
Deep Learning - 5 years
Machine Learning - 5 years
Python - 5 years
TensorFlow - 4 years
Scikit-learn - 4 years
Pandas - 3 years
PyTorch - 3 years

Availability

Part-time

Preferred Environment

Bash, Vim Text Editor, Visual Studio, Jupyter Notebook, Git, Slack, Linux

The most amazing...

...thing I've worked on used deep learning for identifying a specific type of brain tumor from slide images that currently cannot be detected visually.

Work Experience

Machine Learning Engineer

2019 - 2020

Freelance (India)

Worked remotely with three startups to build an ML-based MVP.
Collaborated with a group of medical researchers to develop a deep learning-based brain tumor detection and classification algorithm from histopathological data. The research was unprecedented and extends current medical science.
Worked with a London-based company for building automated meter reading using deep learning OCR and deployed it with embedded systems.
Jointly worked with a US-based dental startup to build a teeth whiteness estimator from teeth images using deep learning and computer vision.

Technologies: TensorFlow, PyTorch, Python

Machine Learning Engineer

2018 - 2019

Pathway AI (San Jose, USA)

Contributed from the start to the build of a successful MVP.
Implemented distance estimation and camera calibration with advanced curved lane detection using OpenCV in Python and C++.
Developed a weather classifier using TensorFlow, converted it to CoreML, and deployed it in an iOS app with Swift.

Technologies: Swift, OpenCV, C++, TensorFlow, Python

Research Assistant

2017 - 2018

Fraunhofer Research Institute (Germany)

Developed real-time traffic sign recognition with over 98% accuracy for a BMWi autonomous driving project.
Performed research and developed a Bayesian deep learning framework for uncertainty estimation in semantic scene segmentation for self-driving cars.
Extended state-of-the-art semantic segmentation models to perform in real-time without loss of performance.

Technologies: C++, Python, TensorFlow

Experience

Real-time Semantic Segmentation

I optimized a state-of-the-art deep CNN network to perform real-time predictions at 20 FPS on an NVIDIA V100 GPU with a loss of only 1% performance accuracy. The project was developed using TensorFlow and Python.

Traffic Sign Recognition

I developed a real-time CNN model that was performing predictions with over 0.98 AUROC for an imbalance dataset of 30:1 data classes. I also developed a customized loss function and a better metric to evaluate the imbalance dataset.

Artificial Traffic Sign Image Generation

I trained a CycleGAN to generate realistic-looking traffic signs from computer-generated images of traffic signs. The GAN learned to add a custom background and create high-resolution training data with enough variations providing synthetic data.

Uncertainty Estimation using Bayesian Deep Learning

http://publica.fraunhofer.de/dokumente/N-484660.html

This is my master's thesis. I applied a Bayesian deep learning framework to semantic scenes for autonomous driving to estimate pixel-wise estimation uncertainties. I also modified a state-of-the-art model for Bayesian settings and it was tested on a real autonomous car.

Brain Tumor Detection and Classification

I developed a deep learning CNN-based brain tumor detection and classification algorithm from histopathological data (slide images). Currently, medical science is not aware of visual features that can be used to detect this sub-category of tumor. I also developed a model that provides predictions as well as heatmap visualization of images to explain parts of images that triggered predictions.

Meter Reading OCR

I developed a TensorFlow embedded based real-time OCR to detect digits and to read and process digital and analog meter readings. It was trained with open-source similar data and fine-tuned for this application to mitigate the short-comings of limited data.

Education

2015 - 2018

Master of Science Degree in Autonomous Systems

Bonn Rhein Sieg University of Applied Sciences - Bonn, Germany

2011 - 2015

Bachelor of Engineering Degree in Mechatronics

Gujarat Technological University - Gujarat, India

Skills

Libraries/APIs

TensorFlow, PyTorch, Keras, Scikit-learn, Fast.ai, NumPy, OpenCV, Pandas

Tools

Git, Slack, Visual Studio, Vim Text Editor

Languages

Python, C++, Bash, Swift

Platforms

Linux, Jupyter Notebook

Frameworks

Django

Other

Computer Vision, Deep Learning, Machine Learning, Robot Operating System (ROS)

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.

Share your needs

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

Choose your talent

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

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