Sandeep Kumar, Ph.D.
Associate Professor,Department of Electrical Engineering, School of Artificial Intelligence (joint appointment),Bharti School of Telecommunication Technology and Management (affiliate)Indian Institute of Technology Delhi (IIT Delhi).Phone: (+91)-11-2659-8734 Office: Block II, 413 B.Sandeep Kumar is an associate professor in the Department of Electrical Engineering and School of Artificial Intelligence (ScAI) at the Indian Institute of Technology Delhi (IIT Delhi). He heads the Machine Intelligence Signals and Networks (MISN) lab at IIT Delhi. His current research focuses on graphical models, manifolds, semi-supervised learning, and large-scale optimization. He received his M.Tech and Ph.D. degrees from the Indian Institute of Technology (IIT), Kanpur, and completed his postdoc from the Hong Kong University of Science and Technology.
Research: I supervise the Machine Intelligence Signals and Networks (MISN) Lab at IIT Delhi. The overarching theme of our research efforts lies in the coherent interaction between optimization, machine learning, signal processing, and graphical models. Specific problems of interest include:
Graph Compression
Machine Learning over non-Euclidean Spaces
Fair Federated Personalized Graph Neural Networks
Applications of Graph ML in Earth science, Neuroscience, Wireless communication, and Circuit Design
AI for Tribal Languages
Our Paper "BiMol-Diff: A Unified Diffusion Framework for Molecular Generation and Captioning" got accepted to ACL Main conference, 2026.
● Our paper "Enhancing depression diagnosis with augmented brain signal driven decorrelated graph neural networks" got accepted to [Nature Communications Medicine] .Congratulations to Jyotismita Barman and Mohammad Yusuf.
● Our paper "Fast and Scalable Hashing-Based Universal Graph Coarsening" got accepted to [IEEE Transactions on Pattern Analysis and Machine Intelligence(TPAMI)] .Congratulations to Mohit Kataria and Nikita Malik.
● Our paper "Graph Machine Learning Framework for Gravimetry Data Analysis and Cavity Detection" got accepted to [IEEE Transactions on Artificial Intelligence] .Congratulations to Gourab Panda, Vipul Kumar Singh and Anil Kumar.
● Our paper "GraphGini: Fostering Individual and Group Fairness in Graph Neural Networks" got accepted to [Transactions on Machine Learning Research] .Congratulations to Anuj Kumar Sirohi and Anjali Gupta.
● Our paper "CoarseFuse: Graph-Coarsening-Based Multi-Atlas Functional Connectivity Fusion for Autism Spectrum Disorder Diagnosis" got accepted to [IEEE Journal of Biomedical and Health Informatics (JBHI)] .Congratulations to Ekta Srivastava and Siddhant Ujjain.
● Our paper "Leap of FAITH from GNN-to-MLP: Fairness Aware Inference via DisTillation of GrapH Knowledge " got accepted to [AAAI 2026] .Congratulations to Vipul Kumar Singh and Jyotismita Barman.
● Our paper "GlucoGrapher: Sensor-Aware CGM Fusion with Mask-Aware Meta-Ensembles for Predicting Carbohydrate Caloric Ratio from Postprandial Glucose" got accepted to [AAAI Bridge Program on AI for Medicine and Healthcare-2026] .Congratulations to Siddhant Ujjain, Pooja Singh, Ekta Srivastava and Ahmad Siraj Hashmi.
● Our paper "REAMAP: A GRAPH-CONSTRAINED SPD CHANGE OPERATOR FOR EEG REACTIVITY WITH CROSS-SESSION AND CROSS-DATASET TRANSFER" got accepted to [IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP-2026)] .Congratulations to Ekta Srivastava, Siddhant Ujjain and Vaibhav Kagathara.
● Our paper "Multidomain Photostimulation EEG Biomarkers: Neural Entrainment Signatures for Differential Dementia Phenotyping" got accepted to [IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP-2026)] .Congratulations to Vaibhav Kagathara, Hitika Gahlot, Siddhant Ujjain, Pooja Singh and Ekta Srivastava.
● Our paper "Geometry-Aligned EEG Scalpograms: Graph Harmonics Enable Accurate, Calibrated Dementia Subtyping" got accepted to [International Joint Conference on Neural Networks (IJCNN 2026), IEEE WCCI 2026] .Congratulations to Siddhant Ujjain, Ahmad Siraj Hashmi, Pooja Singh and Ekta Srivastava.
● Our paper "CONNECT-PD: Early Detection of Parkinson's Disease Using Temporal Connectivity Graphs from Gait Data" got accepted to [International Conference on Pattern Recognition (ICPR)-2026] .Congratulations to Ekta Srivastava and Siddhant Ujjain.
● Our paper "Bonn EEG as Graphs: Structural vs. Recurrence Networks for Tiny-Data Seizure Screening" got accepted to [National Conference on Communications (NCC) 2026] .Congratulations to Ekta Srivastava and Siddhant Ujjain.
● Our paper "GARuD: Guided Alignment of Representations using Distillation for Ultra-Low-Resource Languages" got accepted to [International Joint Conference on Natural Language Processing & Asia-Pacific Chapter of the Association for Computational Linguistics, (AACL-IJCNLP)-2025] .Congratulations to Debarchan Basu, Shashwat Bhardwaj, Vaibhav Sharma and Pooja Singh.
● Our paper "Wideband Reconfigurable Intelligent Surface Unit Cell Utilizing Asymmetric Resonator Geometry for mmWave Communications" got accepted to [IEEE Future Networks World Forum-2025] .Congratulations to Bhishm Tripathi and Shailesh Mani.
● Our paper "IndiHealthBench: Evaluating LLMs for Clinical Translation Across Indian Linguistic Diversity" got accepted to [ 2025, International Conference on Pattern Recognition and Machine Intelligence (PReMI-2025)] .Congratulations to Pooja Singh,Jay Saraf, Aashrith Sharma, Siddhant Ujjain and Paridhi Jain.
● Our paper "EEG-STGNN: Unified Spatial–Temporal Modeling for Alzheimer’s and FTD Detection" got accepted to [International Conference on Computational Data and Social Networks(CSoNet)-2025] .Congratulations to Siddhant Ujjain, Vaibhav Kagathara, Pooja Singh and Ekta Srivastava.
● Congratulations to Pooja on being selected for the NSF-EMBS-Google Sponsored Young Professional NextGen Scholar Recognition.
● A historic and proud moment for MISN Lab Hon'ble President of India, Smt. Droupadi Murmu, addressed and appreciated our pioneering initiative, Adi Vaani, the world's first AI-powered platform for Janjatiya (tribal) languages. [Rashtrapati Bhavan Press Releases][PIB Press Releases]
● It is a proud moment for MISN Lab as we have officially launched the Beta version of Adi Vaani the world's first AI-powered platform dedicated to Janjatiya (tribal) languages. [PIB Press Releases]
● Our paper "Leveraging the Cross-Domain & Cross-Linguistic Corpus for Low Resource NMT: A Case Study On Bhili-Hindi-English Parallel Corpus " got accepted to [Empirical Methods in Natural Language Processing (EMNLP 2025)] .Congratulations to Pooja, Shashwat and Vaibhav.
● Our paper "REFINE: Enabling Efficient and Trustworthy Modeling of Financial Networks via GNN-to-MLP Knowledge Distillation" got accepted to [12th IEEE International Conference on Data Science and Advanced Analytics (DSAA-2025)] .Congratulations to Vipul and Jyotismita.
● Our paper "CrossMed: A Multimodal Cross-Task Benchmark for Compositional Generalization in Medical Imaging" got accepted to [ International Conference on Biomedical and Health Informatics (BHI-2025)] .Congratulations to Pooja and Siddhant.
● Our paper "Coarse-and-learn: Efficient online node labeling " got accepted to [32nd International Conference on Neural Information Processing(ICONIP-2025)] .Congratulations to Subhanu and Manoj
Our paper "Modularity Aided Consistent Attributed Graph Clustering via Coarsening" has been accepted to TMLR 2025. Congratulations to Yukti, Samarth, and Manoj!
Our paper "GOTHAM: Graph Class Incremental Learning Framework under Weak Supervision" has been accepted to TMLR 2025. Congratulations to Aditya!
