Divya Saxena

I am looking for strongly motivated M.Tech., MS (by research) and PhD students. Please email me your CV (subject: "PhD position: your name"), if interested.

Sep 2025	Our paper, Geometry-Consistent 4D Gaussian Splatting for Sparse-Input Dynamic View Synthesis accepted in IEEE Annual Congress on Artificial Intelligence of Things 2025 (IEEE AIoT 2025).
Aug 2025	Our paper, COIN-GNN: Inductive Spatial-Temporal Prediction for Continuous Distribution Shifts via Graph Neural Networks accepted in IEEE Transactions on Knowledge and Data Engineering (TKDE).
Aug 2025	Our paper, Re-GAN: Data-Efficient GANs Training Via Architectural Reconfiguration accepted in IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI).
July 2025	Joined IIT Jodhpur as Assistant Professor.
April 2025	Appointed Associate Editor, IEEE Internet of Things Journal (IoT-J)(April 2025- ).
July 2025	Our paper, Dynamic Generative Adaptation for Data-Efficient GAN was accepted at IJCNN 2025.
Mar 2025	Delivered a tutorial, “Dynamic Architectural Adaptations for Resource-Efficient Generative AI in IoT Systems”, at IEEE PerCom 2025, Washington DC.
July 2025	Our paper, Data-Efficient Alignment in Medical Imaging via Reconfigurable Generative Networks accepted at WACV 2025.
Feb 2025	Invited speaker: “Transforming Computer Vision with Large Language Models”, ATAL FDP sponsored by AICTE, organized by Graphic Era University, India.
Jan 2025	Our paper, LS-GAN: Layer-Specific Pruning for Data-Efficient Generative Adversarial Networks received the Best Paper Award at ISDIA 2025.
Dec 2025	Our paper, Channel Estimation for IRS-aided OTFS System using Dilated Attention GAN accepted in Wiley Transactions on Emerging Telecommunications Technologies.
Dec 2024	Our paper, Enhancing Motion in Text-to-Video Generation with Decomposed Encoding and Conditioning accepted at NeurIPS 2024.
Aug 2024	Delivered a tutorial, “Redefining Generative Modeling in Data-Limited Environments”, at DASFAA 2024, Gifu, Japan.
Aug 2024	Received Best Presentation Runner-up at the PolyU Research Student Conference (PRSC 2024), for the paper “Addressing Multimodel Forgetting in Continual Learning with Growing Model Capacity”.
Oct 2024	Our paper, NASPrecision: Neural Architecture Search-Driven Multi-Stage Learning for Surface Roughness Prediction in Ultra-Precision Machining accepted in Elsevier Expert Systems with Applications (ESWA).
Sep 2024	Elevated to IEEE Senior Member.
2024	Received the Long-Term Service Award, The Hong Kong Polytechnic University, Hong Kong.
Feb 2024	Invited talk, “Adaptive GANs Architectures for Enhanced Data Efficiency", at GLA University, Mathura.

Divya Saxena is an Assistant Professor in the School of Artificial Intelligence and Data Science at the Indian Institute of Technology, Jodhpur (IIT Jodhpur). She received her Ph.D. in Computer Science and Engineering from the Indian Institute of Technology Roorkee (IIT Roorkee) in 2017, and completed her postdoctoral research (2018–2021) at The Hong Kong Polytechnic University (PolyU, QS World Ranking 54, 2025), where she also served as a Research Assistant Professor (2021–2025).

Divya leads research at the frontier of generative AI, developing foundation models that are adaptive, sustainable, and engineered for practical, high-impact deployment.

Dr. Saxena is a Senior Member of IEEE, currently serves as an Associate Editor for the IEEE Internet of Things Journal (IoT-J), and regularly reviews for leading journals and conferences in AI and machine learning. She is committed to developing AI that is robust, accessible, and impactful, and welcomes collaborations across academia and industry.

Divya Saxena’s research centers on advancing the theory and practice of AI, with a focus on developing models and methods that are both fundamentally novel and practically robust. Her work addresses the growing need for AI systems that are scalable, sustainable, and adaptive, bridging foundational innovation with deployment in complex, real-world environments.

Generative AI

Divya’s research in Generative AI centers on building efficient and adaptive generative models—including GANs, diffusion models, and text-to-image generation, that can operate effectively with limited data and adapt rapidly to new domains. Her work advances techniques for dynamic model adaptation, resource-efficient training, and robust performance across diverse data regimes, aiming to make generative modeling more accessible and practical for real-world applications.

Foundation Models

She also focuses on the design, adaptation, and deployment of foundation models, including large language models (LLMs), vision-language, and multimodal models. This involves developing methods for fine-tuning, parameter-efficient adaptation, continual learning, and real-world deployment, with an emphasis on scalable architectures and resource-aware optimization. Her work contributes to advancing foundation model capabilities while ensuring efficient adaptation for domain-specific and interdisciplinary tasks.

AI Applications & Societal Impact

At the heart of Divya’s research is a commitment to translating AI breakthroughs into meaningful societal impact. She collaborates across disciplines to create solutions for healthcare diagnostics, manufacturing quality assurance, food safety, and smart urban environments. By bridging the divide between theoretical research and field-ready solutions, her work strives to create AI technologies that are robust, inclusive, and accessible, benefiting communities and industries worldwide.