Rishav Gupta

All three projects share one technique: ultrasonic sensing with nothing but the phone itself.

• 

  Measuring blood flow through the skin

  Phonotonos recovers arterial blood-flow velocity waveforms with phase-based ultrasonic Doppler on a commodity smartphone, fusing ultrasound with arterial sound and motion sensing to find the artery and reject hand motion. It measures four clinical Doppler indices with accuracy comparable to dedicated medical devices.

  • MobiSys '26 paper
  • MobiCom '26 demo
  • Code and sample data on GitHub

• 

  Reading tissue mechanics with sound

  Body-contact devices such as haptic wristbands, exoskeletons, and prosthetics press on soft tissue that differs from person to person. I estimate tissue relaxation time from acoustic chirps and their echoes using a phone's built-in speaker and microphone, a first step toward devices that tune themselves to each wearer.

  • MobiCom '26 work-in-progress paper

• 

  Securing phone calls with ultrasound

  AI voice cloning lets scammers impersonate trusted people on calls. Ultrasound Watermark binds live call audio to the caller's articulatory motion (mouth, jaw, and cheek movements sensed with the phone's own ultrasound), which makes the watermark hard to fake.

  • EdgeSP '26 paper

I work with Dr. Dong Li at UMBC, together with collaborators at UMass Amherst, physicians at the University of Maryland, and Dr. Janelle Clark in UMBC's Department of Mechanical Engineering.

1. WiPTo appear

   Smartphone-based Acoustic Tissue Characterization for Personalized Body-contact Systems

   Rishav Gupta, Alyssa N. Maguina, Sithumina Weerarathna, Janelle P. Clark, Dong Li

   ACM MobiCom 2026 · 32nd International Conference on Mobile Computing and Networking, work-in-progress track · Austin, TX

   Abstract: acoustic tissue characterization, MobiCom 2026

   Interactive systems such as haptic wristbands maintain sustained body contact, but variations in soft tissue mechanics can cause the same force to feel different across users. Conventional tissue characterization relies on specialized systems such as ultrasound elastography, limiting its use in personalized devices. We take a first step toward smartphone-based acoustic tissue characterization by investigating if a smartphone can estimate tissue relaxation time using its built-in speaker and microphone. The smartphone probes tissue with acoustic chirps and estimates relaxation behavior from tissue echoes. We validate the approach against laboratory indentation, observing a strong correlation between smartphone and mechanical estimates on a silicone phantom (r=0.96) and a weaker positive correlation on the human forearm (r=0.29). These preliminary results support the feasibility of smartphone-based acoustic tissue characterization for tissue-aware body-contact systems.

   BibTeX: acoustic tissue characterization, MobiCom 2026

   @inproceedings{gupta2026tissue,
     title     = {{WiP}: Smartphone-based Acoustic Tissue Characterization
                  for Personalized Body-contact Systems},
     author    = {Gupta, Rishav and Maguina, Alyssa N. and Weerarathna, Sithumina
                  and Clark, Janelle P. and Li, Dong},
     booktitle = {Proceedings of the 32nd Annual International Conference
                  on Mobile Computing and Networking},
     series    = {MobiCom '26},
     year      = {2026},
     publisher = {ACM},
     address   = {Austin, TX, USA},
     note      = {To appear}
   }

2. DemoTo appear

   Phonotonos: Through-Skin Ultrasonic Blood Flow Sensing Using Smartphones

   Rishav Gupta, Shirui Cao, Jie Xiong, Sunghoon Ivan Lee, Jeremy Gummeson, Dong Li

   ACM MobiCom 2026 · 32nd International Conference on Mobile Computing and Networking, demo track · Austin, TX

   • Phonotonos code on GitHub
   Abstract: Phonotonos demo, MobiCom 2026

   Cardiovascular diseases remain a leading cause of death worldwide, motivating accessible tools for vascular health monitoring. This demo presents Phonotonos, a smartphone-based system for through-skin ultrasonic sensing of blood-flow velocity waveforms and Doppler indices. Phonotonos repurposes smartphones for vascular sensing despite their low ultrasound transmit power, coarse spatial resolution, and lack of beamforming. The system combines phase-based ultrasonic Doppler sensing with arterial heart sounds and inertial measurements to localize arteries, mitigate motion interference, and recover blood-flow waveforms. Our demonstration showcases an end-to-end prototype in which users position a smartphone over an artery, receive real-time placement guidance, observe the recovered blood-flow waveform, and obtain clinically relevant Doppler indices. Through this interactive demo, we highlight the feasibility of transforming smartphones into portable and accessible tools for everyday vascular screening and monitoring.

   BibTeX: Phonotonos demo, MobiCom 2026

   @inproceedings{gupta2026phonotonosdemo,
     title     = {Demo: Phonotonos: Through-Skin Ultrasonic Blood Flow Sensing
                  Using Smartphones},
     author    = {Gupta, Rishav and Cao, Shirui and Xiong, Jie and Lee, Sunghoon Ivan
                  and Gummeson, Jeremy and Li, Dong},
     booktitle = {Proceedings of the 32nd Annual International Conference
                  on Mobile Computing and Networking},
     series    = {MobiCom '26},
     year      = {2026},
     publisher = {ACM},
     address   = {Austin, TX, USA},
     note      = {To appear}
   }

3. WorkshopTo appear

   Ultrasound Watermark: Real-time Acoustic Watermarking against Voice Scams on Smartphones

   Renzhi Hao, Shirui Cao, Yizhu Wen, Rishav Gupta, Dharani Nadendla, Mehran Kafai, Hanqing Guo, Dong Li

   ACM/IEEE EdgeSP 2026 · 9th Workshop on Security and Privacy in Edge Computing, co-located with ACM/IEEE SEC 2026 · Santa Clara, CA

   Abstract (excerpt): Ultrasound Watermark, EdgeSP 2026

   In this paper, we present Ultrasound Watermark, a real-time, smartphone-native framework for protecting live phone calls against voice impersonation attacks. Recent advances in text-to-speech (TTS), voice conversion (VC), and other AI-driven speech synthesis attacks enable adversaries to convincingly mimic trusted individuals during phone calls, creating a growing threat of voice scams. To defend, we introduce a real-time acoustic watermarking system that binds speech to a caller's physical articulatory motion. During a call, the caller's smartphone continuously emits ultrasonic signals and captures reflections from mouth, jaw, and cheek movements using only the built-in speaker and microphone. …

   BibTeX: Ultrasound Watermark, EdgeSP 2026

   @inproceedings{hao2026ultrasoundwatermark,
     title     = {Ultrasound Watermark: Real-time Acoustic Watermarking
                  against Voice Scams on Smartphones},
     author    = {Hao, Renzhi and Cao, Shirui and Wen, Yizhu and Gupta, Rishav
                  and Nadendla, Dharani and Kafai, Mehran and Guo, Hanqing
                  and Li, Dong},
     booktitle = {Proceedings of the 9th ACM/IEEE Workshop on Security
                  and Privacy in Edge Computing},
     series    = {EdgeSP '26},
     year      = {2026},
     publisher = {IEEE},
     address   = {Santa Clara, CA, USA},
     note      = {To appear}
   }

4. ConferenceEmerging Ideas

   Phonotonos: Through-Skin Ultrasonic Blood Flow Sensing Using Smartphones

   Shirui Cao, Jie Xiong, Riishav Guptaa, Sunghoon Ivan Lee, Jeremy Gummeson, Dong Li

   ACM MobiSys 2026 · Proceedings of the 24th Annual International Conference on Mobile Systems, Applications and Services, pp. 911–927

   • Read the paper (DOI 10.1145/3745756.3809246)
   • Code and sample data on GitHub
   Abstract: Phonotonos, MobiSys 2026

   Cardiovascular diseases remain the leading cause of death worldwide, and Doppler blood flow indices play a crucial role in their early detection and diagnosis. Existing ultrasound devices, though effective, are costly, bulky, and unsuitable for daily or at-home use. In this work, we introduce Phonotonos, a system that transforms commodity smartphones into portable and accessible tools for through-skin ultrasonic sensing of blood flow velocity. We demonstrate that despite smartphones' inherent limitations—including low ultrasound power, coarse spatial resolution, and lack of beamforming—blood flow velocity waveforms can still be recovered using novel phase-based signal processing, including nonlinearity cancellation and baseline drift removal. We further propose a triple-modality framework that fuses Doppler ultrasound, arterial sound, and IMU for robust artery localization and interfering motion (e.g., involuntary hand motion) rejection. Extensive simulations, phantom experiments, and IRB-approved human studies validate that our proposed system can measure four key Doppler indices (AT, S/D ratio, RI, PI) with accuracy comparable to dedicated medical devices. These results highlight the potential of smartphones to democratize vascular health monitoring and enable continuous cardiovascular screening in everyday environments.

   BibTeX: Phonotonos, MobiSys 2026

   @inproceedings{cao2026phonotonos,
     title     = {[Emerging Ideas] Phonotonos: Through-Skin Ultrasonic
                  Blood Flow Sensing Using Smartphones},
     author    = {Cao, Shirui and Xiong, Jie and Guptaa, Riishav and Lee, Sunghoon Ivan
                  and Gummeson, Jeremy and Li, Dong},
     booktitle = {Proceedings of the 24th Annual International Conference
                  on Mobile Systems, Applications and Services},
     series    = {MobiSys '26},
     pages     = {911--927},
     year      = {2026},
     month     = jun,
     publisher = {ACM},
     doi       = {10.1145/3745756.3809246}
   }

• Graduate Research Assistant

  Future Sensing and Interaction (FSI) Lab, UMBC · Baltimore, MD

  Smartphone ultrasound sensing for cardiovascular health (see Research).

  Jun 2025 – Present 1 year 5 months

• Senior iOS Engineer

  Gojek (GoTo Group) · Bengaluru, India

  Built in-house analytics tooling to replace a third-party SDK, and automated the bi-weekly iOS release pipeline.

  Feb 2022 – Sep 2024 2 years 8 months

• Senior iOS Developer & Team Lead

  Nuclei · Bengaluru, India

  Founding iOS developer of FedMobile for Federal Bank; promoted to lead a team of four iOS developers.

  Jul 2019 – Feb 2022 2 years 8 months

• Android Developer Intern

  Hera doo · Mostar, Bosnia and Herzegovina

  Built an end-to-end Android sales-order application.

  Jun – Jul 2018 2 months

• Ph.D., Computer Science

  University of Maryland, Baltimore County · Advisor: Dr. Dong Li

  Jan 2026 – Present

• M.S., Computer Science

  University of Maryland, Baltimore County · GPA 4.00/4.00

  Jan 2025 – Dec 2025

• B.Tech., Information Technology

  Manipal Institute of Technology · Ranked 8th of 132

  Jun 2015 – May 2019

• COEIT Computing for Social Good Award

  COEIT Research Day, UMBC · 1 of 2 awards across 114 posters

  May 2026

• 1st Place, Maryland NEW VENTURE™ Program Pitch Night

  bwtech@UMBC · for PocketCardio

  May 2026

• Best Poster Award

  CSEE Research Day, UMBC

  Apr 2026

• Best Poster Award

  COEIT Research Day, UMBC

  May 2025

• Chair, Graduate Assistant Advisory Committee (GAAC)

  UMBC Graduate Student Association

  2026 – 2027

• Lead Volunteer

  ACM/IEEE Symposium on Edge Computing (SEC 2025)

  Dec 2025

• HR Head

  ISTE Students’ Chapter, Manipal Institute of Technology

  2017 – 2018

• Co-Event Head (Acumen)

  TechTatva ’17, Manipal Institute of Technology

  Oct 2017

• Management Committee Member

  ISTE Students’ Chapter, Manipal Institute of Technology

  2016 – 2017