About

Rajesh Gupta serves as the Dean of the School of Computing, Information and Data Sciences at UC San Diego, and is the founding director of the Halıcıoğlu Data Science Institute at UC San Diego. He is also a distinguished professor of Computer Science and Engineering at UC San Diego. His research focuses on embedded and cyber-physical systems, with particular emphasis on sensor data organization and its application in optimization and analytics. His notable contributions include the development of SystemC modeling and SPARK parallelizing high-level synthesis, both of which have been incorporated into industrial practice. Gupta has led multiple center-scale efforts on Microelectronics and system design sponsored by NSF and DARPA, and is currently a co-PI on the DARPA/SRC Center on Evolvable Computing, aiming to build a new generation of distributed accelerated computing systems. He has received numerous awards, including the IEEE Computer Society W. Wallace McDowell Award, the NSF CAREER Award, and the Excellence in Stewardship Award from UC San Diego Foundation. Gupta holds the Qualcomm Endowed Chair in Embedded Microsystems at UC San Diego, is an INRIA International Chair at the French research institute in Rennes, and a Distinguished Visiting Professor at IIT Kanpur. He is a Fellow of the IEEE, ACM, and AAAS.

Research topics

• Computer Science
• Artificial Intelligence
• Real-time computing
• Machine Learning
• Data Mining
• Engineering
• Telecommunications
• Electrical engineering
• Computer network
• Embedded system

Selected publications

• Matching Skeleton-based Activity Representations with Heterogeneous Signals for HAR

  2025-05-04 · 1 citations

  articleOpen accessSenior author

  In human activity recognition (HAR), activity labels have typically been encoded in one-hot format, which has a recent shift towards using textual representations to provide contextual knowledge. Here, we argue that HAR should be anchored to physical motion data, as motion forms the basis of activity and applies effectively across sensing systems, whereas text is inherently limited. We propose SKELAR, a novel HAR framework that pretrains activity representations from skeleton data and matches them with heterogeneous HAR signals. Our method addresses two major challenges: (1) capturing core motion knowledge without context-specific details. We achieve this through a self-supervised coarse angle reconstruction task that recovers joint rotation angles, invariant to both users and deployments; (2) adapting the representations to downstream tasks with varying modalities and focuses. To address this, we introduce a self-attention matching module that dynamically prioritizes relevant body parts in a data-driven manner. Given the lack of corresponding labels in existing skeleton data, we establish MASD, a new HAR dataset with IMU, WiFi, and skeleton, collected from 20 subjects performing 27 activities. This is the first broadly applicable HAR dataset with time-synchronized data across three modalities. Experiments show that SKELAR achieves the state-of-the-art performance in both full-shot and few-shot settings. We also demonstrate that SKELAR can effectively leverage synthetic skeleton data to extend its use in scenarios without skeleton collections.

  PublisherDOI

• ZeroHAR: Sensor Context Augments Zero-Shot Wearable Action Recognition

  Proceedings of the AAAI Conference on Artificial Intelligence · 2025-04-11 · 2 citations

  articleOpen access

  Wearable Human Action Recognition (wHAR) uses motion sensor data to identify human movements, which is essential for mobile and wearable devices. However, traditional wHAR systems are only trained on a limited set of activities. Hence, they fail to generalize to diverse human motions, prompting Zero-Shot Learning (ZSL). Existing ZSL methods for wHAR focus solely on augmenting labels, such as representing them as attribute matrices, images, videos, or text. We propose ZeroHAR that enhances ZSL by not just focusing on activity labels, but by augmenting motion data with sensor context features. Our approach incorporates information about the sensor type, the Cartesian axis of the data, and the sensor's body position, providing the model with crucial spatial and biomechanical insights. This helps the model generalize better to new actions. First, we train the model by aligning the latent space of the motion time-series with its corresponding sensor context, while distancing it from unrelated sensor contexts. Finally, we train the model using the target activity descriptions. We tested our method against eight baselines on five benchmark HAR datasets with various sensors, placements, and activities. Our model shows exceptional generalizability across 18 motion time series classification benchmark datasets, outperforming the best baselines by 262% in the zero-shot setting.

  PublisherOA PDFDOI

• AeroSafe: Mobile Indoor Air Purification Using Aerosol Residence Time Analysis and Robotic Cough Emulator Testbed

  2025-05-19

  article

  Indoor air quality plays an essential role in the safety and well-being of occupants, especially in the context of airborne diseases. This paper introduces AeroSafe, a novel approach aimed at enhancing the efficacy of indoor air purification systems through a robotic cough emulator testbed and a digital-twins-based aerosol residence time analysis. Current portable air filters often overlook the concentrations of respiratory aerosols generated by coughs, posing a risk, particularly in high-exposure environments like healthcare facilities and public spaces. To address this gap, we present a robotic dual-agent physical emulator comprising a maneuverable mannequin simulating cough events and a portable air purifier autonomously responding to aerosols. The generated data from this emulator trains a digital twins model, combining a physics-based compartment model with a machine learning approach, using Long Short-Term Memory (LSTM) networks and graph convolution layers. Experimental results demonstrate the model's ability to predict aerosol concentration dynamics with a mean residence time prediction error within 35 seconds. The proposed system's real-time intervention strategies outperform static air filter placement, showcasing its potential in mitigating airborne pathogen risks.

  PublisherDOI

• Networks-on-Chip and High-Level Synthesis: The Power of Abstraction [Special Section on 2025 IEEE Kirchhoff Award]

  IEEE Circuits and Systems Magazine · 2025-01-01

  articleSenior author

  This article gives a perspective on the enduring legacy of Prof. Giovanni (Nanni) De Micheli and his fundamental contributions on raising the level of abstraction in the design of Systems-on-chip, focusing specifically on his seminal work on Networks-on-chip and high-level synthesis.

  PublisherDOI

• Pedagogical Leadership and the Implementation of Project-Based Curriculum: Impact on Students' Problem-Solving Abilities in Australia

  JMPI Jurnal Manajemen Pendidikan dan Pemikiran Islam · 2025-03-29

  articleOpen access

  This study explores the impact of pedagogical leadership and the implementation of project-based learning (PBL) on students' problem-solving abilities in Australia. In the context of modern education marked by the challenges of globalization and technological advancement, the PBL approach has proven to enhance student engagement in the learning process through real-world projects that promote critical thinking and problem-solving skills. Pedagogical leadership, which includes support from school management and a culture of collaboration, plays a crucial role in fostering innovation in the implementation of PBL. This study employs qualitative methods, including in-depth interviews, observations, and document analysis, to gain a comprehensive understanding of how PBL is applied and its effects on students' problem-solving abilities. The research findings indicate that supportive pedagogical leadership, which encourages innovation and collaboration among teachers as well as parental involvement, can enhance student motivation and create a conducive learning environment. However, challenges such as limited teacher training and difficulties in integrating projects with national curriculum standards pose obstacles to the effective implementation of PBL. The study concludes that successful PBL implementation requires support from strong school leadership, provision of resources, and training for teachers. These findings have significant implications for education policy in Australia, particularly in designing a more relevant curriculum that prepares students to face real-world challenges.

  PublisherOA PDFDOI

• Orthogonal Calibration for Asynchronous Federated Learning

  ArXiv.org · 2025-02-21

  preprintOpen access

  Asynchronous federated learning mitigates the inefficiency of conventional synchronous aggregation by integrating updates as they arrive and adjusting their influence based on staleness. Due to asynchrony and data heterogeneity, learning objectives at the global and local levels are inherently inconsistent -- global optimization trajectories may conflict with ongoing local updates. Existing asynchronous methods simply distribute the latest global weights to clients, which can overwrite local progress and cause model drift. In this paper, we propose OrthoFL, an orthogonal calibration framework that decouples global and local learning progress and adjusts global shifts to minimize interference before merging them into local models. In OrthoFL, clients and the server maintain separate model weights. Upon receiving an update, the server aggregates it into the global weights via a moving average. For client weights, the server computes the global weight shift accumulated during the client's delay and removes the components aligned with the direction of the received update. The resulting parameters lie in a subspace orthogonal to the client update and preserve the maximal information from the global progress. The calibrated global shift is then merged into the client weights for further training. Extensive experiments show that OrthoFL improves accuracy by 9.6% and achieves a 12$\times$ speedup compared to synchronous methods. Moreover, it consistently outperforms state-of-the-art asynchronous baselines under various delay patterns and heterogeneity scenarios.

  PublisherOA PDFDOI

• Contextual Inference From Sparse Shopping Transactions Based on Motif Patterns

  IEEE Transactions on Knowledge and Data Engineering · 2025-01-10 · 1 citations

  article

  Inferring contextual information such as demographics from historical transactions is valuable to public agencies and businesses. Existing methods are data-hungry and do not work well when the available records of transactions are sparse. We consider here specifically inference of demographic information using limited historical grocery transactions from a few random trips that a typical business or public service organization may see. We propose a novel method called <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DemoMotif</small> to build a network model from heterogeneous data and identify subgraph patterns (i.e., motifs) that enable us to infer demographic attributes. We then design a novel motif context selection algorithm to find specific node combinations significant to certain demographic groups. Finally, we learn representations of households using these selected motif instances as context, and employ a standard classifier (e.g., SVM) for inference. For evaluation purposes, we use three real-world consumer datasets, spanning different regions and time periods in the U.S. We evaluate the framework for predicting three attributes: ethnicity, seniority of household heads, and presence of children. Extensive experiments and case studies demonstrate that <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DemoMotif</small> is capable of inferring household demographics using only a small number (e.g., fewer than 10) of random grocery trips, significantly outperforming the state-of-the-art.

  PublisherDOI

• Approximate Lie Symmetry Analysis: Exact and Approximate Solutions of the Singularly Perturbed Generalized Hodgkin–Huxley Equation

  International Journal of Applied and Computational Mathematics · 2025-03-14 · 2 citations

  articleSenior author
  PublisherDOI

• Matching Skeleton-based Activity Representations with Heterogeneous Signals for HAR

  ArXiv.org · 2025-03-17

  preprintOpen accessSenior author

  In human activity recognition (HAR), activity labels have typically been encoded in one-hot format, which has a recent shift towards using textual representations to provide contextual knowledge. Here, we argue that HAR should be anchored to physical motion data, as motion forms the basis of activity and applies effectively across sensing systems, whereas text is inherently limited. We propose SKELAR, a novel HAR framework that pretrains activity representations from skeleton data and matches them with heterogeneous HAR signals. Our method addresses two major challenges: (1) capturing core motion knowledge without context-specific details. We achieve this through a self-supervised coarse angle reconstruction task that recovers joint rotation angles, invariant to both users and deployments; (2) adapting the representations to downstream tasks with varying modalities and focuses. To address this, we introduce a self-attention matching module that dynamically prioritizes relevant body parts in a data-driven manner. Given the lack of corresponding labels in existing skeleton data, we establish MASD, a new HAR dataset with IMU, WiFi, and skeleton, collected from 20 subjects performing 27 activities. This is the first broadly applicable HAR dataset with time-synchronized data across three modalities. Experiments show that SKELAR achieves the state-of-the-art performance in both full-shot and few-shot settings. We also demonstrate that SKELAR can effectively leverage synthetic skeleton data to extend its use in scenarios without skeleton collections.

  PublisherOA PDFDOI

• Reliability-Based Design of Water Distribution Network

  Lecture notes in civil engineering · 2025-01-01

  book-chapterSenior author
  PublisherDOI

Recent grants

• CPS:Small:Collaborative Research:Localization and System Services for SpatioTemporal Actions in Cyber-Physical Systems

  NSF · $250k · 2009–2012

• Collaborative Research: Variability-Aware Software for Efficient Computing with Nanoscale Devices

  NSF · $4.0M · 2010–2017

• Modeling and Optimization of Thermal and Energy Efficient Processing in Multi-Core System-Chips

  NSF · $275k · 2007–2010

• CSR:Small:Collaborative Research:EDS: Systems and Algorithmic Support for Managing Complexity in Sensorized Distributed Systems

  NSF · $200k · 2015–2018

• Constrained Power and Performance Optimization for Embedded Systems

  NSF · $157k · 2003–2007

Frequent coauthors

• Luca Benini

  ETH Zurich

  51 shared

• Yuvraj Agarwal

  Carnegie Mellon University

  51 shared

• Abbas Rahimi

  IBM Research - Zurich

  49 shared

• Sandeep K. Shukla

  Indian Institute of Technology Kanpur

  45 shared

• Dezhi Hong

  Amazon (United States)

  41 shared

• Bharathan Balaji

  32 shared

• Nikil Dutt

  28 shared

• Giovanni De Micheli

  École Polytechnique Fédérale de Lausanne

  22 shared

Education

• Ph. D., Electrical Engineering

  Stanford University

  1994

• M. S., EECS

  University of California, Berkeley

  1986

• B. Tech., Electrical Engineering

  Indian Institute of Technology Kanpur

  1984

Awards & honors

• Excellence in Stewardship Award from UC San Diego Foundation
• IEEE Computer Society W. Wallace McDowell Award
• Distinguished Alumnus Award from IIT Kanpur
• IEEE TCCPS Distinguished Leadership Award
• National Science Foundation CAREER Award