Ashwin Phadke

Ashwin Phadke

Computer Vision | Deep Learning

Biography

Ashwin worked as a AI and Deep Learning Engineer helping apply computer vision applications in the field of mobility, healthcare and industrial QC/QA. With around 2 years of total experience in computer vision and deep learning his research interests also include artificial neural netwoks and natural language processing.

He has been a speaker at leading tech conferences and meetups. He also serves as a mentor for incubated startups by working with startup founders to upskill their workforce in the field of deep learning, computer vision. He has also trained 500+ students by speaking at various mentoring sessions and seminars in colleges.

Interests

  • Artificial Intelligence
  • Deep Learning
  • Computer and Machine vision
  • Natural Language Processing
  • Information Retrieval

Education

  • BEng in Electronics and Telecommunications, 2018

    P.E.S Modern College of Engineering

News and Upcoming Talks

Know more about my upcoming or previously given talks and workshops also news that I’d like to share.

Analyzing Model Performance using Tensorflow Profiler

Model performance is always a important aspect while training or tuning deep learning models. It is often left to experience or referred to certain previous benchmarks on different data and made analysis upon.

Experience

 
 
 
 
 

AI and Deep Learning Engineer

Cynapto Technologies

Jun 2019 – Oct 2019
  • Managed a team of 4 AI Engineers
  • Develop computer vision solutions for automotive, industrial QA and securtity domains.
  • Gather, develop and deploy computer vision solutions on-prem, cloud and edge devices.
  • Develop image processing and video analysis solutions for multi object detection and tracking, facial recognition, heat maps , segmentation , instance differentiation.
  • Worked on optimizing algorithm performance over cpu.
 
 
 
 
 

Junior Engineer - AI and Deep Learning

Cynapto Technologies

Jun 2019 – Aug 2018
• Design, Implemented and supervised the development of computer vision and machine learning algorithms for products using neural networks. • Improve computational and algorithmic compatibility for algorithms, NVidia GPU’s(1080) for multi-processing and threading, async calls for better efficiency. • Develop image processing and video analysis solutions for multi object detection and tracking, facial recognition, heat maps , segmentation , instance differentiation. • Implement supervised and unsupervised machine learning algorithms for analytics and statistics using regression, classification, visualization techniques. • Worked on tuning, training and modelling algorithms on self made and open source datasets accounting to 300GB+.
 
 
 
 
 

Deep learning and Computer vision Mentor

CLASS Private Limited

Mar 2019 – May 2019
  • Mentored a NASSCOM incubated startup in starting up their deep learning and computer vision journey by setting up and managing a team of interns and training them in deep learning concepts. • Trained and helped develop deep learning based application for their app based product and running inferences on device and cloud. • As a social responsibility to guide students and startups in their AI journey

Accomplish­ments

Get a peek at all my accomplishments, certificates, awards and recognitions.

Convolutional Neural Networks in Tensorflow

See certificate

Introduction to Tensorflow

See certificate

Machine Learning with Python

Deep Learning Fundamentals
See certificate

Neural Networks and Deep Learnig

See certificate

Blog

Blog posts that I have authored, sometimes alo featuring guest authors but such posts are explicitly mentioned.

Load TensorFlow Models Using OpenCV

Background : It is always a daunting task with Tensorflow sessions and standard handling of a typical Tensorflow model when you want to run inference. However, if you are an experienced developer you may also quickly go through these steps because you are already aware about how to use Tensorflow to run inference on your model.

Build your own layers for deep learning models using TensorFlow 2.0 and Python

Background During a very recent webinar where I was a speaker for the topic Deep learning with TensorFlow I repeatedly was asked a question regarding how would one really define their own layers, parameter and how they work so as to watch it do the magic while showing them some notebooks that had parameters to the layers that we regularly use.

Projects

.js-id-Deep-Learning

Computer Vision Playground

A computer vision playground to try and test end to end(test to deploy) computer vision pipeline. We are looking for open source enthusiasts to help advance the project, to know more click on contribute.

Object Detection

Object detection usin OPenCV’s dnn module a mobilenet model trianed on caffe.

Smart Anti Breach

An app to secure your lockers, with fencing location and multi role access web.

Face Recognition using Python and OpenCV Haarcascades

Implementing simple facial recognition using Haarcascades and OpenCV

Optical Flow using OpenCV and Python

An implementation of optical flow tracker using lucas-k : calcOpticalFlowPyrLK() method in OpenCV.

Canny Edge Detection using OpenCV and C++

implementing canny edge detection from OpenCV with noise reduction, intensity gradient of image, non-maxima suppression and hysterisis thresholding

Find Dominant Color from an image using OpenCV and C++

Finding dominant color in an image with the help of quantization and eigne value.