Utkarsh Nath

Utkarsh Nath

I am a PhD Candidate in the School of Computing and Augmented Intelligence, at Arizona State University. I work as a research assistant in the Geometric Media Lab, under the guidance of Dr. Pavan Turaga. My research spans Computer Vision and Generative AI, with recent focus on 3D asset generation, text-to-3D systems, and retrieval-augmented generation (RAG) for enterprise applications.
Connect at unath@asu.edu

News

Recent Industrial Research Experience

LinkedIn
(May 2025 - Aug 2025)
PhD GenAI Research Intern ⚙️ Patent Pending
  • Proposed Agentic Memory RAG (AMRAG): a domain-aware RAG framework combining semantic and relational memory within a schema-bounded planning paradigm for global query resolution.
  • Designed a dual-memory indexing strategy using embedding-based semantic memory and relational storage, eliminating expensive LLM calls at ingestion time.
  • Developed a retrieval-time compiler that transforms each query into a deterministic DAG with three bounded operations: Semantic Search, Relational Query, and LLM Synthesis.
  • Outperformed state-of-the-art RAG pipelines on LinkedIn Recruiter data, achieving 33% higher accuracy, 2.6× faster indexing, and up to 150× lower cost.

Selected Publications

DecompDreamer
DecompDreamer: Advancing Structured 3D Asset Generation with Multi-Object Decomposition and Gaussian Splatting
Utkarsh Nath, Rajeev Goel, Rahul Khurana, Kyle Min, Mark Ollila, Pavan K. Turaga, Varun Jampani, Tejaswi Gowda
In Submission ICLR 2026
Paper / Project
Deep Geometric Moments
Deep Geometric Moments Promote Shape Consistency in Text-to-3D Generation
Utkarsh Nath, Rajeev Goel, Eun Som Jeon, Changhoon Kim, Kyle Min, Yezhou Yang, Yingzhen Yang, Pavan Turaga
WACV 2025
Paper / Project
Poly-INR
Polynomial Implicit Neural Framework for Promoting Shape Awareness in Generative Models
Utkarsh Nath, Rajhans Singh, Ankita Singh, Kuldeep Kulkarni, Pavan Turaga
International Journal of Computer Vision (IJCV), 2024
Paper
Poly-INR
Learning Low-Rank Features for Thorax Disease Classification
Yancheng Wang*, Rajeev Goel*, Utkarsh Nath, AC Silva, Teresa Wu, Yingzhen Yang
NeurIPS 2024
Paper
RNAS-CL
RNAS-CL: Robust Neural Architecture Search by Cross-Layer Knowledge Distillation
Utkarsh Nath, Yancheng Wang, Pavan Turaga, Yingzhen Yang
International Journal of Computer Vision (IJCV), 2024
UAI 2024
Paper
RNAS-CL
Adjoined Networks: A Training Paradigm with Applications to Network Compression
Utkarsh Nath, Shrinu Kushagra, Yingzhen Yang
AAAI Spring Symposium, 2022
Paper

Work Experience

Samsung Research
(July 2018 - July 2019)
Software Engineer
  • Led a team of three to build a mobile application to interact and control internal functioning of Samsung Smart TV through wireless (wifi-direct) and wired connection.
  • Features of application involved controlling factory settings, fetching serial logs, running internal tests and fixing them.
Coding Blocks
(Aug 2017 - July 2019)
Algorithm Instructor
  • Conducted Launchpad course for C++: Data Structures, Algorithms, Object Oriented Programming.
  • Taught batch of 60 students at a time: includes preparing assignments, quizzes, doubt-solving sessions.
Google Summer of Code / FOSSASIA
(May 2017 - Aug 2017)
Student Developer
  • Worked on Open-Event project, which aims to develop automated tool for creation of app and website for conferences. Part of the team responsible for frontend development and designing of the tool.
  • Used Semantic UI components to build responsive UI, EmberJS in back-end and GitHub for version control.

Education

Arizona State University
(Expected May 2026)
Doctor of Philosophy, Computer Science

New York University
(May 2021)
Master of Science, Computer Science

Delhi Technological University
(May 2018)
Bachelor of Technology, Information Technology

Service

Reviewing

  • NeurIPS 2024, 2025
  • ICLR 2025, 2026
  • WACV 2025
  • ICML 2025
  • IEEE Transactions on Information Forensics and Security
  • IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

Teaching Assistant

  • Data Structures and Algorithms (CSE 310, ASU)
  • Foundation of Machine Learning (CSE 475, ASU)
  • Statistical Machine Learning (CSE 575, ASU)
  • Introduction to Programming (CS 1114, NYU)

Patents