Search Results - chaitali+chakrabarti

10 Results

Sort By:

Full-stack Obfuscation Tool to Mitigate Neural Architecture Stealing

The architecture information of a Deep Neural Network (DNN) model is considered a valuable, sensitive piece of property for a company. Knowledge of a DNN’s exact architecture allows any adversary to build a substitute model and use this substitute model to launch devastating adversarial attacks. Side-channel based DNN architecture stealing can...

Published: 2/14/2024 | Inventor(s): Jingtao Li, Chaitali Chakrabarti, Deliang Fan, Adnan Siraj Rakin

Keywords(s): Artificial Intelligence, Cyber Security, Defense Applications, Machine Learning, Neural Computing, PS-Computing and Information Technology

Category(s): Computing & Information Technology, Physical Science, Intelligence & Security

Hierarchical Coarse-Grain Sparsity for Deep Neural Networks

Background Recurrent neural networks (RNNs) that enable accurate automatic speech recognition (ASR) are large in size and have long short-term memory (LSTM) capabilities. Due to the large size of these networks, most speech recognition tasks are performed in the cloud servers, which requires constant internet connection, introduces privacy concerns,...

Published: 2/23/2023 | Inventor(s): Jae-Sun Seo, Deepak Kadetotad, Chaitali Chakrabarti, Visar Berisha

Keywords(s):

Category(s): Physical Science, Computing & Information Technology, Wireless & Networking

Binary Neural Network for Improved Accuracy and Defense Against Bit-Flip Attacks

Recently, Deep Neural Networks (DNNs) have been deployed in many safety-critical applications. The security of DNN models can be compromised by adversarial input examples, where the adversary maliciously crafts and adds input noise to fool a DNN model. The perturbation of model parameters (e.g., weight) is another security concern, one that relates...

Published: 2/24/2023 | Inventor(s): Deliang Fan, Adnan Siraj Rakin, Li Yang, Chaitali Chakrabarti, Yu Cao, Jae-Sun Seo, Jingtao Li

Keywords(s): Algorithm Development, Artificial Intelligence, Cyber Security, Defense Applications, Machine Learning, Neural Computing

Category(s): Physical Science, Intelligence & Security, Wireless & Networking, Computing & Information Technology

User-Space Emulation Framework for Domain-Specific Soc Design

This technology uses Linux-based emulation framework to provide an ecosystem for hardware-software co-design of Domain-specific SoCs (DSSoCs) and enable their rapid evaluation during the pre-silicon design phase. Background: As technology scaling becomes a challenge, System-on-chip architects are exploring DSSoCs to effectively balance performance...

Published: 8/1/2024 | Inventor(s): Umit Ogras, Radu Marculescu, Ali Akoglu, Chaitali Chakrabarti, Daniel Bliss, Samet Egemen Arda, Anderson Luiz Sartor, Nirmal Kumbhare, Anish Krishnakumar, Joshua Mack, Ahmet Alper Goksoy, Sumit Kumar Mandal

Keywords(s):

Category(s): Technology Classifications > Software & Information Technology > Computer-Aided Design & Engineering

Hierarchical and Lightweight Imitation Learning for Power Management of Embedded Socs

This technology involves an imitation learning based power management system for use in embedded systems-on-chips (SoCs). Background: As technology advances, systems-on-chips (SoCs) need to be able to operate more efficiently to keep up with multiple rapidly adapting technologies. Many SoCs struggle with dynamic power management and need to be able...

Keywords(s):

Category(s): Technology Classifications > Software & Information Technology > Computer-Aided Design & Engineering

Runtime Task Scheduling using Imitation Learning for Heterogeneous Many-Core Systems;

This technology involves task scheduling to optimally utilize the processing elements at runtime using imitation learning. Background: Harvesting the full potential of many-core systems-on-chips is difficult. The goals are to minimize execution time, power dissipation, and energy consumption. This technology intends to solve this problem through assigning...

Keywords(s):

Category(s): Technology Classifications > Software & Information Technology > Computer-Aided Design & Engineering

PMU-Embedded Analytics for Online Event Detection and Classification in Power Grids

The energy requirement of the world is increasing exponentially every year. With the power demand sites concentrated in the region with the highest population density, energy production sites must be located remotely. This isolation requires lengthy and complicated grids to be designed and manufactured to transmit power over long distances. Grid failures...

Published: 7/30/2024 | Inventor(s): Payman Dehghanian, Shiyuan Wang, Li Li

Keywords(s):

Category(s): Technology Classifications > Industrial or Consumer Tech > Energy Infrastructure and Environment

Coarse-Grain Memory Sparsification for Small-Footprint Deep Neural Networks

Recent breakthroughs in deep neural networks (DNNs) have led to improvements in state-of-the-art speech applications. Conventional DNNs have hundreds or thousands of neurons in each layer, which require a large amount of memory to store the connections between neurons. Implementing these networks in hardware requires a large memory and high computation...

Published: 2/23/2023 | Inventor(s): Jae-Sun Seo, Chaitali Chakrabarti, Sairam Arunachalam, Deepak Kadetotad

Keywords(s):

Category(s): Computing & Information Technology, Wireless & Networking

Distributed Feedback Fiber Laser Pumped by Multimode Diode Lasers

Researchers at the University of Arizona College of Optical Sciences and Canada's Carleton University recently succeeded in fabricating high reflectivity (>99%) Fiber Bragg Gratings (FBGs) into phosphate glass fibers using UV light and a phase mask technique. With this new manufacturing technique, fiber lasers can be created as single monolithic...

Published: 4/3/2023 | Inventor(s): Nasser Peyghambarian, Axel Schülzgen, Li Li, Jacques Albert

Keywords(s):

Category(s): Technology Classifications > Imaging & Optics, Technology Classifications > Imaging & Optics > Lasers & Other Sources

All-Fiber Multicore Fiber Laser Devices

Researchers at the University of Arizona have developed a compact, power-scalable, alignment-free fiber laser that takes a new approach to stabilizing phase-locked operation of multi-core fiber lasers by completely removing all free space optical components. This high-brightness, all-fiber laser package uses passive optical fiber spliced at both ends...

Published: 4/3/2023 | Inventor(s): Nasser Peyghambarian, Axel Schülzgen, Li Li

Keywords(s):

Category(s): Technology Classifications > Imaging & Optics, Technology Classifications > Imaging & Optics > Fiber Optics, Technology Classifications > Imaging & Optics > Lasers & Other Sources