Akankshya Patra

Welcome to a Computer Science Programmer's bio!

About Photo

I was born and brought up in Rourkela, India. I am a young engineering in making, currently pursuing Masters of Science in Computer Science at University of Southern California, Los Angeles.

I am super active and a curious cat. My academic interests lie in the field of Machine Learning. Computational Linguistics to be more precise! You will know why...

And my non-academic passion would be a summation of "I am a dancer at heart, an avid quizzer, and an art enthusiast."

Download printable Resume! Or you can still have a look at it.

Professional info

As an undergraduate in Computer Science and Engineering, I have the fundamental knowledge of Data Structures, Algorithm Analysis. I have a deep understanding of a broad set of algorithms and applied maths, probability, and statistics. I have good analytical and problem-solving skills. I am a quick learner and always try to adapt and improvise my thoughts and application into a problem. Apart from being curious and inquisitive, I have a tender nature of experimenting things. Through continuous learning, I have had the opportunity to gain the depth of knowledge in the realm of Machine Learning.

Programming Skills

•Programming: Python, MATLAB, C++, Java(basic)

• General application: Tensorflow, Pytorch, NLTK, OpenCV, Arduino Uno, SQL

• Familiar with OS/Environments: Windows, Linux(Ubuntu)

Work Experience

MACHINE TRANSLATION OF LOW RESOURCE LANGUAGES USING NEURAL NETWORKS
May-July 2018,MITACS Globalink Research Internship, University of New Brunswick, Canada

The idea was to preserve the indegenous language- Mi'kmaq (native to New Brunswick). Due to lack of monolingual documents and corpora, the challenge was real. Using LSTM as the neural network, I implemented Cross-lingual word Embedding from a rich-resource language such as English, and hence the translation task!​​

DEPTH IMAGE INPAINTING USING DEEP IMAGE PRIOR
December 2017,Research Intern at Indian Institute of Technology Kharagpur, India.

After the invention of Depth Images, the most certain way to prevent bad quality image is to reconstruct the Depth Images. One of the major technique is Image inpainting. Using CNN and different Priors (Deep image, GMRF, Superpixel) along with the corresponding RGB image, we could revive the lost portion of the depth images.​​

PATTERN RECOGNITION
May 2017 - June 2017, Summer Research Intern at Indian Statistical Institute, Bangalore Bangalore, India.

Using the University Student dataset, Implemented Minimum distance classifier, KNN, Decision tree to find similar patterns of daily cycle. Further implementation of Naive Bayes and Neural Networks to provie the University with a relaxed timetable.

PERFORMANCE ANALYSIS OF LOSS-LESS METHODS OF DATA COMPRESSION.
December 2016, Research Intern at National Institute of Technology, Rourkela, India.

Implemented Huffman encoding, Run-length encoding, and the Lempel-Ziv encoding for text data compression. Analyzed on different text sets.

Projects

VIDEO CAPTIONING USING HYBRIDISED DEEP LEARNING METHODS
November 2018 - April 2019: Undergraduate Thesis, NIT Rourkela, India

Using LSTM models with feedback and attention mechanism along with Tag based model, improved on the result of getting captions from a Video. How humans can describe a video is mimicked by the Computers on your laps.

VISUAL RECOGNITION TO TEXT
Feb 2018 - April 2018: Product Development Laboratory, NIT Rourkela, India

How captions generated from an Image descriptor give a better result in terms of Syntax and semantics. Based on the obtained result, improved upon the LSTM network to give a semantically meaningful caption

MACHINE TRANSLATION FROM ENGLISH TO ODIA(NATIVE LANGUAGE) USING NLP.
Sep.2017-Nov.2017: Product Development Laboratory, NIT Rourkela, India

This is one of the very first works to develop a Machine translator system for Odia. Availability of no benchmark daatset was the biggest challenge. Therfore using Syntax and Rules, implemented Machine Translation task. The words were mapped from a English o Odia dictionary which was extracted by Web-scraping using Python(Beautiful Soup). Using Stanford Parser and NLTK, to implement reordering of sentences. Tense reconstruction and mapping to Odia words from the corpus was implemented.

CENTRAL DATABASE SYSTEM
Oct. 2016-Dec. 2016: Cyborg, NIT, Rourkela, India

A central database which stores information of the members of the club to facilitate easy task of exchanging and sharing electronic stuffs. Implemented front-end coding using HTML5, CSS, and Javascript.

LINE FOLLOWER BOT
Oct. 2015-Nov. 2015: Cyborg, National Institute of Technology, Rourkela Rourkela, India

Constructed a line follower bot that was required to follow a white line on a black path. Implemented embedded coding in Arduino Uno platform.

Education

2019-2021
    Master of Science in Computer Science, University of Southern California
    Gpa: N/A
2015-2019
    B.Tech in Computer Science and Engineering, National Institute Of Technology, Rourkela
    Gpa: 9.00/10
2015
    All India Senior School Certificate Examination, Deepika English Medium School, Rourkela
    Percentage:95.8
2013
    Indian Certificate of Secondary Education, St. Joseph’s Convent School, Rourkela
    Percentage: 95.6

Achievements

    Awarded the MITACS Globalink Scholarship-2018 to pursue a Research Internship
    2nd in Indian Hip-hop Dance Championship and qualified for World-HHI.
    Best Academic performance award from SAIL, India in CBSE class-XII board examination.
    Academic Excellence Award-2015 from Youth Movement Federation of India in CBSE class-XII board examination.
    Academic Excellence Award-2013 from Youth Movement Federation of India in ICSE class-X board examination.
    Qualified state level in National Talent Scholarship Examination, India.
    Qualified in Junior Mathematics Olympiad-2011 and certified from Odisha Mathematical Society and Institute of Mathematics and Applications, India.

A bit more about me...

Dancing

SYNERGY: CORE-MEMBER & VICE-PRESIDENT, KALABHUMI: CORE-MEMBER AND MENTOR

  • Member of SnM Crew. Secured 2nd position in Indian Hip-hop Dance Championship and qualified for World HHI.
  • Invited and Performed at India International Dance Festival-2017.
  • Invited and Performed at Chausathi Yogini International Dance Festival-2016.
  • "Creative" prize at hip-hop dance, NIT Warangal.
  • Winners at Hip-Hop dance, Terpsichore, BITS Hyderabad.
  • Runners-up in Choreonite at IIT Kharagpur Spring Fest.
  • Awarded Nritya-Visharad in Bharatanatyam-2014.

Technical

CYBORG:ROBOTICS AND AUTOMATION, NIT ROURKELA- CORE-MEMBER

  • Participation in Robotics events across the country.
  • Programming and coding.
  • Coordinated in the creation of club website and the Cyborg Central Database System.
  • Conducted online coding contest and workshops.

INQUIZZITIVE: QUIZZING CLUB, NIT ROURKELA- MEMBER

  • Active participation in quiz events across India.
  • In-house quizzing sessions.

testimonial

Leave a message

Web e-mail

Contact:+91-9937174403