Roopesh 
 
Roopesh N 
Nallakshyam  27, Borewell road 
Nallurhalli, Whitefield 
Bangalore - 560066 
 
+91 7729024706, 6302339501 
roopeshnallakshyam@gmail.com
 

Objective  To explore and improve my skills by accepting the challenges and 

finding solutions for them in Computer Vision and Machine learning 
areas. 

ㅡ   

Skills  Image Processing, Computer Vision, Machine Learning, Deep 

Learning, OpenCV, TensorFlow, C++, Python, and Git. 

ㅡ   

Experience 
InnovatorsBay Technologies Pvt. Ltd/ ​Data Evangelist - 
Computer Vision Specialist. 
March 2020 - Present, Bangalore. 
Here I worked on various use-cases for CCTV footage surveillance in 
the JSW steel plant CCTV Security System. 
 
1. PPE-Helmet Detection​: If any person inside the plant found 
without wearing personal protection helmet on their head, 
then we raise alert, with severity high. 
2. Social Distancing(Covid-19): ​The AI models detect all the 
persons in the frames and draw a bounding box on them, 
calculate the distance between each person in pixels, and 
mapping it to real-world meters, based on the threshold of 
the distance between each person, we decide whether social 
distancing is maintained or not.  
3. Movement of people in Restricted ad Secluded Areas: ​There 
will be some areas marked as restricted/secluded areas 
where people movement is not allowed, our algorithm will 
keep on checking those areas and if there are any people 
movement found there it raises alert to control room. 
4. Crowd Detection in the plant: ​It will detect the crowd of 
people more than 5 staying for more than a certain time 
duration then raise alert that people are there as a crowd for 
more than the time fixed. 
5. Vehicle Speed Detection​: Detect the type of vehicle car, bike, 
truck, etc. and then calculate the speed of that particular 
 vehicle using object tracker, If the speed exceeds the speed 
limit set for each type of vehicle then raise the alert. 
6. Confided Spaces​: There will be some regions in the frame 
that are marked with polygons which are called confined 
spaces, whereinto those regions strictly no people are 
allowed, if any person entering that region we raise the alert. 
7. Vehicle Speed + Vehicle Number Plate​: This is the extension 
vehicle speed alert, when a vehicle exceeds the speed limit 
then we detect and recognize that vehicle number plate 
number and save it to the database, also raise the alert that a 
vehicle with vehicle number exceeds the speed limit. 
 
Cognitive Machines Software Solutions/ ​Associate Computer 
Vision and Machine Learning Engineer. 
September 2019 - February 2020, Bangalore. 
Here I worked on various projects with Computer Vision and 
Machine Learning Technologies: 
1. Distracted Driving:​ When a car driver using mobile while 
driving, our application runs in the background and notifies 
him that you are distracting from driving. 
2. Floord: ​Laying a floor tile texture on the indoor floors, in this, 
I worked on the perspective transformation to give users the 
best experience.
3. Intuision​: Lane deviation, when a car driver is crossing the 
lane marks entering to another lane we notify him with our 
application. 
4. Pixm:​ Companies logo detection on webpages such as 
Google, Facebook, Paypal, etc. 
5. Sports Analytics:​ Player detection and saying for which team 
he belongs to by using player wearing jersey color and 
pattern. 

Cognitive Machines Software Solutions / ​Computer Vision and 

Machine Learning Intern. 
February​ 2019 - ​September​ 2019, ​Bangalore. 
In this intern also I worked on projects using Computer Vision and 
Machine Learning Technologies: 
 
1. ULC Robotics:​ We implemented an application for counting 
the number of pipes jointed in the water pipeline tunnel and 
finding where are the tapholes in it. 
2. Zetwerk: ​Bill of Materials table detection and extracting the 
material details from the table from the engineering 
drawings. 
 
 
 MateLabs Innovations private Ltd / ​Machine Learning Intern 
August 2018 - November 2018, Bangalore 
Here I worked on Deep Learning ImageNet Architectures for 
Image classification using Convolutional neural networks and 
Testing the MateVerse platform for machine learning 
classification and regression models. 

GeethSoft Solutions And Business Automation / ​Data Science 

Intern 
May 2018 - July 2018, Work from Home 
In this Internship, I have done two tasks one resume classification
and chatbots development for employees and managers in the
company. Resume classification is done with NLP and TF-IDF
algorithms for a little range of resumes and we implemented a
Chabot using RNN and LSTM neural networks.

ㅡ   

Education  Rajiv Gandhi University of Knowledge and Technologies / ​Bachelor 

of Technology. 
June​ 2015 -​ May ​2019, ​Kadapa, Andhra Pradesh. 
I graduated from Electronics and Communication Engineering. 
 
CGPA: 8.16 

Rajiv Gandhi University of Knowledge and Technologies /​XII 

(SENIOR SECONDARY) 
June 2013 - May 2015, Kadapa, Andhra Pradesh. 
I completed my Senior Secondary in Mathematics, Physics, Chemistry 
stream. 
CGPA: 9.2 

ZPHS, PACHIKAPALLAM, AND CHITTOOR. /​X ( SECONDARY) 

June 2013, Chittoor, Andhra Pradesh. 
 
CGPA: 10.0 
 
 
 

ㅡ   

Courses  ● Machine Learning​ by Andrew NG in Coursera online 

● Convolutional Neural Networks for visual recognition
(CS231N) ​by Stanford University 
● Fundamentals of Computer Vision​ by Mubarak Shah
● Image Understanding with TensorFlow on GCP (Google
Cloud Platform)​ ​In Coursera Online. 
  

Projects  ● Object detection and recognition device for blind people

(Feb 2018-June 2018)
We tried to implement a device to recognize the objects
and human faces, animals, etc. for blind people, using
computer vision and machine learning technologies.

● Motion-Detection and Tracking (Jun 2018-Jul 2018)

In a video or any live streaming video, this will detect moving
objects such as cars, bikes, person, etc. by marking them
with rectangular boxes​. 
 
https://github.com/Roopesh-Nallakshyam/Motion-D
etection-and-Tracking

● Resume classification for particular Job (May 2018-Jun

2018) 
I have collected some resumes who are willing to apply for
a job, with nltk (Natural Language toolkit) and TF_IDF
algorithm I extracted keywords such as technical skills,
projects, etc, and based on that words we classified the
person’s role.
https://github.com/Roopesh-Nallakshyam/Resume-c
lassification-for-particular-job 
 
● Image Classification with K-Nearest Neighbor Algorithm
(March 2018-Apr2018)
We have collected image data set of 5 classes of images
with each of 50 examples we extracted features of these
images by using SIFT (Scale-invariant feature transform)
and saved them into CSV files as mini train and mini-test.
We collected the applied K-Nearest neighbor algorithm
which finds a given new example is closer to the given
dataset and then it classifies that into the nearest class of
image it gives that class as output.

https://github.com/Roopesh-Nallakshyam/Image-classific
ation-with-KNN-algorithm

● Transporter Bot for eYantra Robotics Competition 2k17

(Sep 2017-Jan 2018)
 Transporter Bot is an autonomous robot used for
agriculture purposes we have done this project using the
FIREBIRD V ATMEGA2560 platform as a part of eYantra
Robotics competition 2k17 held by IIT Bombay. This is a
prototype model to reduce the manpower to load the crop
bags into a truck in the large agriculture field.

● IoT based household electrical appliances automation

(Dec 2017 - Dec 2017)​
In this project, we control household electrical appliances
such as TV, lights, fans, etc. using IoT. Here we use Arduino
based Ethernet shield for generating HTML for users to
control which device to be on /off.

  ● Panorama (Image Stitching). (June 2017-Nov 2017)

Panorama stitching of two or more images by overlaying the 
same features of images. In other words, combining images 
based on the same features into a single image using SIFT 
algorithm and opencv. We extract the features of images and 
compare this image features with another then the detected 
same features will be appended together and forms a new 
image.