S
Soumik Sarkar
Researcher at Iowa State University
Publications - 288
Citations - 7113
Soumik Sarkar is an academic researcher from Iowa State University. The author has contributed to research in topics: Deep learning & Computer science. The author has an hindex of 31, co-authored 258 publications receiving 4542 citations. Previous affiliations of Soumik Sarkar include Indian Institute of Science & Raytheon.
Papers
More filters
Posted Content
Root-cause analysis for time-series anomalies via spatiotemporal causal graphical modeling.
TL;DR: A new data-driven framework for root-cause analysis based on a spatiotemporal feature extraction scheme for multivariate time series built on the concept of symbolic dynamics for discovering and representing causal interactions among subsystems of a complex system is presented.
Proceedings ArticleDOI
Battery-Free Camera Occupancy Detection System
TL;DR: In this article, a human occupancy detection system using battery-free cameras and a deep learning model implemented on a low-cost hub to detect human presence is presented. But, the system is limited to the presence of a single person.
Proceedings ArticleDOI
Symbolic identification for anomaly detection in aircraft gas turbine engines
TL;DR: This paper presents a robust and computationally inexpensive technique of fault detection in aircraft gas-turbine engines, based on a recently developed statistical pattern recognition tool.
Proceedings Article
Attribute-Controlled Traffic Data Augmentation Using Conditional Generative Models
TL;DR: An approach where attribute-based generative models conditioned on the time-of-day labels to reconstruct semantically valid transformed versions of the original data to circumvent the challenge of resource-intensive and expensive collection and annotation of traffic data.
Posted Content
Usefulness of interpretability methods to explain deep learning based plant stress phenotyping
Koushik Nagasubramanian,Asheesh K. Singh,Arti Singh,Soumik Sarkar,Baskar Ganapathysubramanian +4 more
TL;DR: This work trains a DenseNet-121 network for the classification of eight different soybean stresses (biotic and abiotic) and trains some of the most popular interpretability methods, including Saliency Maps, SmoothGrad, Guided Backpropogation, Deep Taylor Decomposition, Integrated Gradients, Layer-wise Relevance Propagation and Gradient times Input, for interpreting the deep learning model.