Sponsored Research

Selected Current Research Topics

Corruption-Resistant Data Analysis and Feature Extraction Based on L1-norm Projections (Funded by AFOSR YIP)

Traditional data analysis and signal processing is based on squared residual errors and projections. This is motivated by statistical foundations (scaled sum of squares estimates variance) and ease in optimization methods (e.g., least-squares, SVD). However, learning metrics based on the sum of squares (L2-norm) benefits highly deviating / peripheral points (sensitivity against outliers). In this research direction, we explore theory and develop algorithms for data analysis and feature extraction based on the sum of absolute values (L1-norm) of data projections. While L1-norm minimization has been used for decades in signal processing and data science as a way to promote sparsity, in this research direction we do something different: we use L1-norm maximization as a means to promote balanced focus across all processed data points and, thus, achieve robustness against faulty data and adversarial contaminations.

Dynamic and Robust Tensor Data Analysis (Funded by NSF OAC and AFOSR)

Tensors are multi-way arrays that can generalize matrices to higher orders. Tensor modeling and processing are ubiquitous in modern data analysis and machine learning, as they capture inherent multi-linear structures within the data. For example, tensors are used for modeling multi-relational graphs, multi-spectral images/videos, wireless spectrum utilization maps, and even the parameters of neural networks. While matrix analysis has been thoroughly studied for a very long time, there is still much room (and need) for research in tensor analysis. Primarily funded by NSF and AFOSR, in this research direction we develop new methods for dynamic and robust analysis of tensor data.

Incremental and Continual Learning Based on Neural-Network Parameter Rank Updates (New Topic; Funded by AFRL)

Typically, the more the trainable parameters, the higher the network capacity and the higher its ability to represent more complex models. However, this is only attainable when there is a correspondingly high number of data to train on. Otherwise, a large network can overfit the limited training examples and fail to properly generalize to unseen data. In this work, we focus on adjusting the network capacity to the available data by tuning the tensor rank of its parameters. Then, we apply this idea to incremental learning (increase capacity as more data become available) and continual learning (learn new tasks, in new multi-linear subspaces of the network parameters).

Object Detection/Tracking in Multimodal Aerial Imagery and Neural Network Fusion for Generic Multimodal Learning (New Topic; Funded by NGA)

In this research direction, we develop neural network architectures that are specifically designed for object detection and tracking in aerial imagery (can handle rotated objects, small scarce/objects, and limited training examples). In addition, we explore novel methods for in-the-network fusion that allow for the use of multi-modal aerial imagery for object detection. In a new thread of this project, we develop generic methods for mid-level network fusion applicable to a wide range of data modalities and problems.


Summary: Total: ~$2,082,403 since 2016 (~$2M external, ~$872K as PI, ~$1.21M as Co-PI).

Active Grants:

Title: Theory and Efficient Algorithms for Dynamic and Robust L1-Norm Analysis of Tensor Data.

Agency: U.S. Air Force Office of Scientific Research (AFOSR).

Period: January 2020 - January 2023.

Role: Sole PI.

AFOSR Young Investigator Program Award.

Title: Target Detection/Tracking and Activity Recognition from Multimodal Data.

Agency: National Geospatial-Intelligence Agency.

Period: September 2019 - September 2024.

Role: Equal-effort co-PI.

Title: Collaborative Research: CDS&E: Theoretical Foundations and Algorithms for L1-Norm-Based Reliable Multi-Modal Data Analysis.

Agency: U.S. National Science Foundation (NSF).

Period: September 2018 - August 2021.

Role: PI.

Title: Efficient Radar Imaging and Machine Learning for Concealed Object Detection.

Agency: NYSTAR / UR CoE in Data Science.

Period: October 2021 - June 2022.

Role: Sole PI.

Title: Continual and Incremental Learning with Tensor-Factorized Neural Networks.

Agency: U.S. Air Force Research Laboratory (AFRL).

Period: September-December 2021.

Role: Sole PI.

Completed Grants:

Title: Data-Driven Adaptive Learning for Video Analytics.

Agency: U.S. Air Force Office of Scientific Research (AFOSR).

Period: February 2018 - February 2021.

Role: Co-PI.

Title: Efficient Methods for Dynamic and Robust Analysis of Tensors.

Agency: U.S. Air Force Research Laboratory (AFRL).

Period: November-December 2020.

Role: Sole PI.

Title: Development and Testing of Robust Algorithms for Real-Time Recognition of Complex Gait Patterns from Wearable Sensor Data.

Sponsor: Kate Gleason College of Engineering (KGCOE), RIT.

Period: May 2018 - December 2019.

Role: PI.

Title: Methods for Corruption-Resistant Analysis of Tensor Data.

Agency: U.S. Air Force Research Laboratory, Information Directorate (AFRL/RI).

Period: September 2018 - October 2018.

Role: Sole PI.

Title: Practical L1-Norm Principal Component Analysis: Tools for Reliable Data Analytics.

Sponsor: Office of Vice President of Research (OVPR), RIT.

Period: April 2016 - March 2017.

Role: Sole PI.

Title: Distributed Self-Localization of Wireless-Node Squads in Hostile Environments.

Sponsor: Harris Corporation (now part of L3-Harris).

Period: November 2016 - June 2017.

Role: PI.

Selected Research Honors

  • Young Investigator Program (YIP) Award, Air Force Office of Scientific Research (AFOSR), 2020.

  • Exemplary Performance in Research, Kate Gleason College of Engineering, RIT, 2019, for the research proposals submitted in 2018.

  • Exemplary Performance in Research, Kate Gleason College of Engineering, RIT, 2018, for the research proposals submitted in 2017.

  • Runner-up Poster Award, IEEE Western New York Image and Signal Processing Workshop, 2018, for the paper "Gait recognition based on tensor analysis of acceleration data from wearable sensors."

  • Student Travel Grant Award, SPIE Defense and Commercial Sensing, 2017, for the paper "Adaptive sparse-binary waveform design for all-spectrum channelization."

  • Student Travel Grant Award, SPIE Defense, Security, and Sensing, 2014, for the paper "Direction finding with L1-norm subspaces."

  • Best Paper Award in Physical Layer Communications and Signal Processing, IEEE/VTS/EURASIP International Symposium on Wireless Communication Systems, 2013, for the paper "Some options for L1-subspace signal processing."