Research Statement

I am a computer science student with prior research experience in Human-Computer Interaction, Natural Language Processing and Machine Learning. I have worked in leading research labs in HCI with specializations ranging from embodied conversational agents for education to novel 3D printing interfaces. I have collaborated with graduate and post-graduate researchers at leading institutions like CMU and MIT in finding solutions to problems like virtual collabortative child agents and creating novel interfaces to explore 3D printing models. The fundamental aim of my research is to design sociable, intelligent and interactive technologies that amplify human ability to realize imagination, motivate self-reliance and enhance creative expression.

Research Experience

Visiting Student

Jan 2019 - Present
HCI Engineering Group, CSAIL, Massachusetts Institute of Technology Prof. Stefanie Mueller
  • Ecosytem for Slicing: Designed a system that reverses the traditional 3D printing workflow and allows exploration of multiple prototyping options simultaneously.
  • Using a popular slicing software, created a parallel slicing system that instantiated in the cloud, an online storage system to reuse previously processed results, and a desktop and web-based UI to explore the results.
  • Identification using 3D Print Markers: Collaborated in developing a novel method to identify 3D printed objects by utilizing the unique patterns created by modifying the slicing parameters during the 3D printing process.
  • Developed an Acoustic detection system that detects the sound created by scratching the surface of 3D models, analyzes its frequency spectrum to extract MFCC features and classifies the correct class of slicing setting.

Ungergraduate Research Assistant

June - August 2018
Human-Computer Interaction Institute, Carnegie Mellon University Prof. Justine Cassell
  • Sensing Curiosity in Play and Responding: Developed an intelligent virtual child as a peer collaborator to study behavior and evoke curiosity in a multiparty game called Outbreak.
  • Designed a model to dynamically extract optimal audio signal from multiple sources, perform Automatic Speech Recognition, and convert the unstructured speech to a knowledge representation.
  • Created a Restricted Wizard-of-Oz system to efficiently detect and annotate conversational behaviors shown by children while playing the game. The data generated from the study was used to analyse the frequency of behaviors and the various dependencies among them.

Ungergraduate Research Assistant

June - July, 2017
Language Technologies Research Center, IIIT Hyderabad, India Prof. Manish Shrivastava
  • Semantic Analysis of Humor in English Language: Created a dataset of 800,000 humor and non-humor sentences extracted from social media websites and other web-based sources.
  • Created an Attention-based LSTM model to classify sentences as humorous or non-humorous.
  • Denormalized Text Generator: Designed an algorithm to generate code-mixed data from a parallel corpora of English and Hindi using Linguistic Features.
  • Developed a text-generator that used frequency as a metric to generate bilingual text with most common words of each language.

Ungergraduate Research Assistant

January, 2017 - December, 2018
Manipal Institute of Technology, India Prof. P.B. Shanthi
  • Cervical Cancer Detection System: Explored various Image Enhancement techniques like including Canny-Edge, Sobel and Fuzzy Normalization to augment target features in segmented cell images.
  • Used Deep Convolutional Networks to classify processed cell images. Explored Transfer Learning as a method for improving accuracy.
  • Shunya: Lightweight Handwritten Digit Recognizer: Using Kirsch Directional features of an image as input, and a Convolutional Network as a classifier, created a lightweight system for handwritten digit recognition. Achieved 98.8% accuracy in MNIST using a fraction of compute power to train as compared to traditional CNNs.

Publications

  • Microscopic Uterine Cervix Cell Classification for Malignancy Detection
  • Shanthi P.B., Faruqi F., Hareesha K.S., Kudva R.
    Asian Pacific Journal of Cancer Prevention
  • Is This a Joke?: A Large Humor Classification Dataset
  • Faruqi.F, Shrivastava M.
    International Conference on Natural Language Processing 2018

    Presentations

    MIT-IBM Watson AI Lab Workshop

    May, 2019
    Poster Presentation Cambridge, MA
    • DeepFab: Learning to Predict Optimal Slicing Parameters for Additive Manufacturing
    • IntelliMaterial: Learning to Predict Optimal Slicing Parameters for New Materials in Additive Manufacturing

    Collective Intelligence Workshop at MIT Quest for Intelligence

    September, 2019
    Poster Presentation Cambridge, MA
    • IntelliSlice: Reusing and Sharing slicing results

    Technical Skills

    Programming Languages


    • Adept in C, Java, Python and Javascript
    • Working Knowledge of C#, C++, QML

    Database


    • SQL Server, Oracle Express

    Sofware, IDE


    • MATLAB, Eclipse, Android Studio, Visual Studio, Cura, Adobe Suite, Processing

    Relevant Coursework

    Data Structures & Algorithms, Object Oriented Programming, Artificial Intelligence, Machine Learning, Graph Theory, Automata Theory, Database Systems, Microprocessors, Sofware Engineering, Software Engineering, Computer Networks

    Leadership/ Extra-Curricular Activities

    • Campus Ambassador for Google Applied CS with Android course facilitated by Google.
    • Organized Awareness Campaigns, Door to Door visits and cleanliness campaigns as part of Swachh Bharat Abhiyaan which recently received Gates Foundation Award.
    • Computer Science Instructor for Helpage Youth Foundation.
    • Former Core Committee member of ACM Chapter Manipal.
    • Recieved Bronze medal at ASISC State level swimming competition.
    • Published Essays in Weekly Newspapers and School Yearbooks.