Hi, I am Nakul Garg, a fourth year Ph.D student at the University of Maryland College Park working at iCoSMoS Lab with Prof. Nirupam Roy.
Research interests: Low-power sensing, Mobile Computing, Embedded AI, Wireless Networking.
Lab : IRB 3245
Office : IRB 2116

Publications

  • SPiDR: Ultra-low-power Acoustic Spatial Sensing for Micro-robot Navigation
    Yang Bai, Nakul Garg (co-primary), Nirupam Roy
    Mobisys 2022 Best paper award
    [pdf] [slides] [teaser]

  • Poster: Ultra-low-power Acoustic Imaging
    Yang Bai, Nakul Garg (co-primary), Nirupam Roy
    Mobisys 2022
    [pdf]

  • VoiceFind: Noise-Resilient Speech Recovery in Commodity Headphones
    Irtaza Shahid, Yang Bai, Nakul Garg, Nirupam Roy
    IASA Workshop, Mobisys 2022
    [pdf]

  • Owlet: Enabling Spatial Information in Ubiquitous Acoustic Devices
    Nakul Garg, Yang Bai (co-primary), Nirupam Roy
    Mobisys 2021
    [pdf] [slides] [talk] [teaser]

  • Demo: Microstructure-guided Spatial Sensing for Low-power IoT
    Nakul Garg, Yang Bai (co-primary), Nirupam Roy
    Mobisys 2021 Best demo award
    [pdf] [demo]

  • Enabling Self-defense in Small Drones
    Nakul Garg, Nirupam Roy
    HotMobile 2020
    [pdf] [talk] [poster]

  • Poster: Acoustic Sensing for Detecting Projectile Attacks on Small Drones
    Nakul Garg, Nirupam Roy
    HotMobile 2020
    [pdf]

  • Poster: DRIZY- Collaborative Driver Assistance Over Wireless Networks
    Nakul Garg, Ishani Janveja, Divyansh Malhotra, Chetan Chawla, Pulkit Gupta, Harshil Bansal, Aakanksha Chowdhery, Prerana Mukherjee, Brejesh Lall
    MobiCom 2017
    [pdf] [poster]

Awards and Honors

  • Best paper award at ACM MobiSys, 2022
  • Three minute Thesis (3MT) Award, 2022
  • Owlet featured on cover of ACM GetMobile June edition, 2021
  • Best demo award at ACM MobiSys, 2021
  • University of Maryland CS summer research fellowship, 2021
  • ACM HotMobile 2020 student travel award, 2020
  • University of Maryland graduate school dean’s fellowship, 2019
  • Outstanding student volunteer award by IEEE delhi section, 2018
  • President of India award at World food hackathon, 2017
  • ACM MobiCom 2017 student travel award, 2017
  • Winning team, Celestini project India, 2017
  • Winning team, eYantra - national robotics competition, IIT Bombay, 2017
  • Second prize, eYantra - national robotics competition, IIT Bombay, 2016
  • First prize in state level CBSE science exhibition, 2012
  • First prize in annual school science exhibition, 2009, 2010 and 2011

Past projects (before joining UMD)

    Air Pollution Sensor Network
    We set up a sensor network of air pollution monitoring devices in 200 Delhi Buses. Our sensor nodes can measure PM 1.0, 2.5 & 10.0, temperature, humidity, pressure, number of vehicles around, location, and speed of vehicle. This network can give a better insight of the air pollution in city than the existing 7 stationary air quality measuring stations. Advised by Professor Rijurekha Sen.
    [Times of India] [Hindustan Times]

    Near Sensor Machine Learning
    Implementing neural networks with binary and ternary bit widths on low energy devices like Arduino and Raspberry Pi 3, which cannot handle the other memory-hungry and compute-intensive deep neural networks. Professor Siddharth Joshi is designing a camera sensor which can perform spatial convolutions in analog domain, and I am investigating what hardware-software combination can give the best trade-off between accuracy, latency, cost, and energy by interfacing the new camera sensor with different processors such as ARM Cortex A53 quad-core on Raspberry PI 3B, ARM Cortex M3 on Arduino Due, and others. Advised by Professor Rijurekha Sen.

    AQI Prediction | Celestini Project 2018
    I was appreciated for my work ethics by the Celestini program directors and was appointed as the Technical Advisor for the 2018 Celestini Project. I mentored a team to build a smartphone application that allows users to upload an input image of the sky horizon taken from their smartphone camera and predict the air quality particulate matter indicator, PM2.5 concentration. The team ended up winning the Celestini Prize and was recognised by a myriad of media sources.
    [Financial Express] [Hindustan Times] [NDTV] [Business Standard] [First Post] [India Today]
    [Video] [Website] [Play Store]

    Drizy : Collaborative Driving Assistance | Celestini Project 2017
    As a part of the Celestini Project, sponsored by Google and Marconi Society and mentored by Dr. Aakanksha Chowdhery, we developed Drizy: Drive Easy, a collaborative driver assistance system that can alert drivers of impending collisions. Our system was prototyped to infer two types of collisions: vehicle-to-vehicle based on GPS sensor data of vehicles uploaded to the cloud and vehicle-to-pedestrian based on video feed from vehicles’ dashboard camera. We were awarded the Paul Baron Young Scholars Celestini Prize for our novel solution and also had the opportunity of presenting our work at the renowned ACM MobiCom 2017, USA.
    [Video] [Website]