# About Thomas Carr, Ph.D.

Thomas Carr is an AI researcher and software engineer based in Charlotte, North Carolina. He earned his **Ph.D. in Computing and Information Systems** from the **University of North Carolina at Charlotte** in **May 2026**, with a dissertation titled *Preserving User Privacy on Skeleton-Based Motion Data*.

## Research Focus

Thomas's research investigates privacy preservation for skeleton-based motion data in virtual and augmented reality settings. Although skeleton data appears anonymous, it encodes personally identifiable information through anthropometric structure and motion style, enabling re-identification. His dissertation contributes a family of attack and defense models that together characterize the privacy–utility trade-off:

- **Linkage Attack Neural Network (LAN)** — a Siamese classifier that re-identifies individuals across skeleton sequences.
- **Privacy-centric Deep Motion Retargeting (PMR)** — adversarially trained retargeting that suppresses identifying features while preserving action content.
- **Explanation-based anonymization** — explainable-AI techniques that localize and mask privacy-sensitive joints.
- **DisentangledTMR** — a factorized-transformer retargeting architecture with explicit identity/action disentanglement.
- **MIRAGE** — a causal, streaming autoregressive anonymizer that operates frame-by-frame with no target skeleton required.

## Current Role

**AI & Software Engineer, Incerta Intelligence** (July 2025 – present). Works on defense contracts focused on multimodal fusion for explainable and auditable decision support. Builds AI systems for mission-critical applications with emphasis on transparency and accountability.

## Education

- **Ph.D., Computing and Information Systems** — UNC Charlotte, 2023 – May 2026. Dissertation: *Preserving User Privacy on Skeleton-Based Motion Data*.
- **M.S., Computer Science** — UNC Charlotte, 2022.
- **B.S., Computer Science** — UNC Charlotte, 2019 – 2021.

## Selected Prior Experience

- **Founding AI & Software Engineer, ACR Technologies** (Dec 2024 – Oct 2025) — built an AI-enhanced intraoperative neural-monitoring communications platform.
- **Graduate Research Assistant, UNC Charlotte** (Jan 2023 – 2025) — published at leading AI/CV venues on privacy-preserving motion analysis and bias/fairness in ML.
- **Full-Stack Developer / Systems Engineer, MDcentric Technology** (May 2021 – Dec 2022) — built a real-time asset-tracking system for 30,000+ devices.
- **Founder / Lead Developer, ViBot** (Mar 2020 – Feb 2024) — a Discord platform serving 125,000 daily users, with a 98%-accuracy ML duplicate-account detector.

## Expertise

Privacy-preserving machine learning, skeleton-based motion data, motion retargeting, virtual-reality privacy, ethical machine learning, explainable AI, deep learning, computer vision, multimodal fusion, full-stack development.

## Identifiers

- ORCID: [0009-0006-6039-0209](https://orcid.org/0009-0006-6039-0209)
- Google Scholar: [scholar.google.com/citations?user=a1uc2zEAAAAJ](https://scholar.google.com/citations?hl=en&user=a1uc2zEAAAAJ)
- GitHub: [github.com/thomasc33](https://github.com/thomasc33)
- LinkedIn: [linkedin.com/in/thomasc33](https://www.linkedin.com/in/thomasc33/)
- Email: thomas@thomasc.tech


---

# Publications — Thomas Carr, Ph.D.

Complete list of peer-reviewed papers, dissertation, and ongoing work. Author: Thomas Carr ([ORCID 0009-0006-6039-0209](https://orcid.org/0009-0006-6039-0209), [Google Scholar](https://scholar.google.com/citations?hl=en&user=a1uc2zEAAAAJ)).

## Dissertation

### Preserving User Privacy on Skeleton-Based Motion Data (2026)

**Thomas Carr.** Ph.D. Dissertation, University of North Carolina at Charlotte, 2026.

This dissertation investigates privacy preservation for skeleton-based motion data in virtual-reality and related settings. Although skeleton data appears anonymous, it encodes personally identifiable information through anthropometric structure and motion style, enabling re-identification. The work contributes a family of attack and defense models — including the Linkage Attack Neural Network (LAN), the Privacy-centric Deep Motion Retargeting (PMR) model, explanation-based anonymization, factorized-transformer retargeting (DisentangledTMR), and streaming autoregressive anonymization (MIRAGE) — that together characterize the privacy–utility trade-off and advance the state of the art in privacy-preserving motion analysis.

## Peer-Reviewed Publications

### DisentangledTMR: Privacy-Preserving Skeleton Motion Retargeting via Factorized Transformers (2026, Under Review)

**Thomas Carr, Shuhan Yuan, Depeng Xu, Aidong Lu.** ECCV 2026 (Under Review). Project page: [tmr.thomasc.tech](https://tmr.thomasc.tech).

A transformer-based motion-retargeting architecture that achieves privacy through explicit architectural disentanglement. Two encoders with complementary inductive biases — temporal convolutions for action, spatial graph convolutions for identity — feed a factorized decoder that fuses their representations through separate cross-attention streams and adaptive gating. A three-stage training curriculum progressively establishes disentanglement, reconstruction, and end-to-end refinement. On three benchmarks, DisentangledTMR substantially reduces re-identification while preserving action recognition, outperforming single-encoder baselines.

### Privacy-centric Deep Motion Retargeting for Anonymization of Skeleton-Based Motion Visualization (2025)

**Thomas Carr, Depeng Xu, Shuhan Yuan, Aidong Lu.** Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2025. Project page: [pmr.thomasc.tech](https://pmr.thomasc.tech).

Explores motion retargeting as a mitigation for privacy leakage in skeleton data. Proposes a Privacy-centric Deep Motion Retargeting (PMR) model that uses adversarial learning to suppress personally identifiable information while transferring motion from an initial user onto a dummy skeleton. Achieves retargeting quality on par with state-of-the-art models while preventing attacker re-identification.

### AnonVis: A Visualization Tool for Human Motion Anonymization (2025)

**Thomas Carr, Ruby Flanagan, Albert Bastakoti, Depeng Xu, Aidong Lu.** ISMAR 2025 — Demo Track.

Interactive VR visualization tool showcasing the Smart Noise anonymization technique, which uses explainable AI to identify privacy-sensitive joints and apply adaptive noise. The VR demonstration enables side-by-side comparison between original and anonymized motions, making privacy–utility trade-offs tangible. Built on a curated dataset processed through a Blender-to-Unity pipeline.

### Explanation-Based Anonymization Methods for Motion Privacy (2025)

**Thomas Carr, Yaxin Zhao, Depeng Xu, Aidong Lu.** Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD) 2025.

Novel explanation-based methods for motion privacy preservation in skeleton data. Leverages explainable-AI techniques to identify and protect sensitive information while maintaining data utility for motion-analysis tasks.

### A Review of Privacy and Utility in Skeleton-based Data in Virtual Reality Metaverses (2024)

**Thomas Carr, Depeng Xu, Aidong Lu.** 2024 IEEE International Conference on Metaverse Computing, Networking, and Communications (MetaCom). [IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/10740130).

Survey of privacy implications in skeleton-based motion data, focusing on the privacy–utility trade-off, current privacy-preserving techniques, and pose-estimation methods. Discusses state-of-the-art action-recognition applications and the ethical implications of skeleton-data use.

### User Privacy in Skeleton-based Motion Data (2024)

**Thomas Carr, Depeng Xu.** 2024 IEEE International Conference on Big Data (BigData). [IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/9796423).

Proposes the LAN baseline attack model and the PMR adversarial defense model, and extends the framework with explanation-guided joint masking and a transformer-based retargeting model designed for real-time applications via autoregressive decoding.

### Linkage Attack on Skeleton-based Motion Visualization (2023)

**Thomas Carr, Aidong Lu, Depeng Xu.** CIKM 2023. [ACM Digital Library](https://dl.acm.org/doi/10.1145/3583780.3615263).

Introduces the Linkage Attack Neural Network (LAN) — a Siamese-network-based classifier that detects whether a target and reference skeleton belong to the same individual. Also employs classical and deep motion retargeting to cast target skeletons onto dummies for anonymization.

## Ongoing / In Progress

### MIRAGE: Motion Identity Removal via Autoregressive Generative Encoding for Privacy-Preserving Skeleton-based Motion Data (2027 Target)

**Thomas Carr, Depeng Xu, Aidong Lu.** AAAI 2027 (Target). Project page: [mirage.thomasc.tech](https://mirage.thomasc.tech).

A causal autoregressive Transformer for online skeleton anonymization that operates frame-by-frame using only the current frame and a bounded history window, with no target skeleton required. Combines a residual decoder that produces identity-suppressing perturbations while preserving the input coordinate distribution, sliding-window causal attention for constant-memory streaming, and dual-level adversarial–cooperative supervision at both latent and output stages. On NTU RGB+D 60, MIRAGE achieves 89.0% action recognition accuracy with 35.6% re-identification accuracy, competitive with retargeting baselines that require target skeletons, while enabling real-time streaming at 262 FPS with constant memory.


---

# Projects — Thomas Carr, Ph.D.

Selected engineering and research projects.

## ViBot — Discord Bot Platform (2020 – 2024)

Comprehensive Discord bot for community organization and moderation in *Realm of the Mad God*, serving 125,000 daily users at peak. Includes user authentication, moderation, a REST API for external web-app interaction, a basic ML model for alt-account detection (98% accuracy), and NLP / web-scraping pipelines for dynamic data gathering. Stack: Node.js, Discord.js, MySQL, TensorFlow.js, Google Cloud, AWS, Express. [GitHub](https://github.com/thomasc33/vibot)

## ViBot.tech — Web App (Fall 2020)

Web interface for the ViBot platform, with real-time server management and data visualization. Discord OAuth2 authentication, hosted on Google Firebase. Stack: React.js, SSL, Firebase, Discord OAuth2.

## Recycling Classification — Computer Vision Web App (Spring 2021)

Computer-vision application for recycling classification from image/video input, built with PyTorch and FastAI. Uses image processing and data augmentation to improve classification accuracy; React.js frontend on Firebase. Stack: Python, PyTorch, FastAI, React.js, Firebase. [GitHub](https://github.com/thomasc33/recycling-classification) · [Live](https://recycling-classification-6db68.web.app/)

## Blue Bounty — Collaborative Note-Taking App (Fall 2020)

Group-project note-taking app built in React.js with a Node.js/Express backend added mid-project. Stack: React.js, Node.js, Express, MySQL. [GitHub](https://github.com/Ksolh/Blue-bounty) · [Live](https://bluebounty-uncc.firebaseapp.com/)

## Thomasc.tech — Personal Portfolio (Fall 2021 – present)

This site. Single-page React 19 portfolio with MUI 7, Framer Motion, tsParticles, emerald + gold dark glass-morphism aesthetic. Hosted on Firebase. [GitHub](https://github.com/thomasc33/Thomasc.tech) · [Live](https://thomasc.tech)

## Charlotte Research Connect — Research Data Visualization (Fall 2021)

Web app visualizing research data for UNC Charlotte faculty. Stack: React.js, Node.js, Express, MySQL. [GitHub](https://github.com/JokkerBang/Charlotte-Research-Connect) · [Live](https://charlotte-research-connect.web.app/)

## C-Track — Enterprise Asset Tracking (Summer 2021 – Fall 2022)

Real-time asset-tracking system for internal use at MDCentric, tracking 30,000+ devices. React.js frontend, Express backend, MS SQL Server, Microsoft OAuth authentication, IIS hosting, user-permission management. [Frontend](https://github.com/thomasc33/Asset-Tracking-Frontend) · [Backend](https://github.com/thomasc33/Asset-Tracking-Backend)

## Research Project Pages

- **PMR — Privacy-centric Deep Motion Retargeting:** [pmr.thomasc.tech](https://pmr.thomasc.tech)
- **DisentangledTMR — Factorized-Transformer Retargeting:** [tmr.thomasc.tech](https://tmr.thomasc.tech)
- **MIRAGE — Streaming Autoregressive Anonymization:** [mirage.thomasc.tech](https://mirage.thomasc.tech)


---

# Contact — Thomas Carr, Ph.D.

- **Email:** thomas@thomasc.tech
- **LinkedIn:** [linkedin.com/in/thomasc33](https://www.linkedin.com/in/thomasc33/)
- **GitHub:** [github.com/thomasc33](https://github.com/thomasc33)
- **ORCID:** [0009-0006-6039-0209](https://orcid.org/0009-0006-6039-0209)
- **Google Scholar:** [scholar.google.com/citations?user=a1uc2zEAAAAJ](https://scholar.google.com/citations?hl=en&user=a1uc2zEAAAAJ)

The contact form on [thomasc.tech](https://thomasc.tech/#contact) routes to the same email via Formspree.