Showing 1–7 of 7 results for author: Saha, S K

Expert-Agnostic Ultrasound Image Quality Assessment using Deep Variational Clustering

Authors: Deepak Raina, Dimitrios Ntentia, SH Chandrashekhara, Richard Voyles, Subir Kumar Saha

Abstract: Ultrasound imaging is a commonly used modality for several diagnostic and therapeutic procedures. However, the diagnosis by ultrasound relies heavily on the quality of images assessed manually by sonographers, which diminishes the objectivity of the diagnosis and makes it operator-dependent. The supervised learning-based methods for automated quality assessment require manually annotated datasets,… ▽ More Ultrasound imaging is a commonly used modality for several diagnostic and therapeutic procedures. However, the diagnosis by ultrasound relies heavily on the quality of images assessed manually by sonographers, which diminishes the objectivity of the diagnosis and makes it operator-dependent. The supervised learning-based methods for automated quality assessment require manually annotated datasets, which are highly labour-intensive to acquire. These ultrasound images are low in quality and suffer from noisy annotations caused by inter-observer perceptual variations, which hampers learning efficiency. We propose an UnSupervised UltraSound image Quality assessment Network, US2QNet, that eliminates the burden and uncertainty of manual annotations. US2QNet uses the variational autoencoder embedded with the three modules, pre-processing, clustering and post-processing, to jointly enhance, extract, cluster and visualize the quality feature representation of ultrasound images. The pre-processing module uses filtering of images to point the network's attention towards salient quality features, rather than getting distracted by noise. Post-processing is proposed for visualizing the clusters of feature representations in 2D space. We validated the proposed framework for quality assessment of the urinary bladder ultrasound images. The proposed framework achieved 78% accuracy and superior performance to state-of-the-art clustering methods. △ Less

Submitted 6 July, 2023; v1 submitted 5 July, 2023; originally announced July 2023.

Comments: Accepted in IEEE International Conference on Robotics and Automation (ICRA) 2023

Slim U-Net: Efficient Anatomical Feature Preserving U-net Architecture for Ultrasound Image Segmentation

Authors: Deepak Raina, Kashish Verma, SH Chandrashekhara, Subir Kumar Saha

Abstract: We investigate the applicability of U-Net based models for segmenting Urinary Bladder (UB) in male pelvic view UltraSound (US) images. The segmentation of UB in the US image aids radiologists in diagnosing the UB. However, UB in US images has arbitrary shapes, indistinct boundaries and considerably large inter- and intra-subject variability, making segmentation a quite challenging task. Our study… ▽ More We investigate the applicability of U-Net based models for segmenting Urinary Bladder (UB) in male pelvic view UltraSound (US) images. The segmentation of UB in the US image aids radiologists in diagnosing the UB. However, UB in US images has arbitrary shapes, indistinct boundaries and considerably large inter- and intra-subject variability, making segmentation a quite challenging task. Our study of the state-of-the-art (SOTA) segmentation network, U-Net, for the problem reveals that it often fails to capture the salient characteristics of UB due to the varying shape and scales of anatomy in the noisy US image. Also, U-net has an excessive number of trainable parameters, reporting poor computational efficiency during training. We propose a Slim U-Net to address the challenges of UB segmentation. Slim U-Net proposes to efficiently preserve the salient features of UB by reshaping the structure of U-Net using a less number of 2D convolution layers in the contracting path, in order to preserve and impose them on expanding path. To effectively distinguish the blurred boundaries, we propose a novel annotation methodology, which includes the background area of the image at the boundary of a marked region of interest (RoI), thereby steering the model's attention towards boundaries. In addition, we suggested a combination of loss functions for network training in the complex segmentation of UB. The experimental results demonstrate that Slim U-net is statistically superior to U-net for UB segmentation. The Slim U-net further decreases the number of trainable parameters and training time by 54% and 57.7%, respectively, compared to the standard U-Net, without compromising the segmentation accuracy. △ Less

Submitted 22 February, 2023; originally announced February 2023.

Comments: Accepted in 9th ACM International Conference on Biomedical and Bioinformatics Engineering (ICBBE) 2022 http://www.icbbe.com/

Journal ref: Proceedings of the 2022 9th International Conference on Biomedical and Bioinformatics Engineering, pp. 41-48

TrustToken, a Trusted SoC solution for Non-Trusted Intellectual Property (IP)s

Authors: Muhammed Kawser Ahmed, Sujan Kumar Saha, Christophe Bobda

Abstract: Secure and trustworthy execution in heterogeneous SoCs is a major priority in the modern computing system. Security of SoCs mainly addresses two broad layers of trust issues: 1. Protection against hardware security threats(Side-channel, IP Privacy, Cloning, Fault Injection, and Denial of Service); and 2. Protection against malicious software attacks running on SoC processors. To resist malicious s… ▽ More Secure and trustworthy execution in heterogeneous SoCs is a major priority in the modern computing system. Security of SoCs mainly addresses two broad layers of trust issues: 1. Protection against hardware security threats(Side-channel, IP Privacy, Cloning, Fault Injection, and Denial of Service); and 2. Protection against malicious software attacks running on SoC processors. To resist malicious software-level attackers from gaining unauthorized access and compromising security, we propose a root of trust-based trusted execution mechanism \textbf{\textit{(named as \textbf{TrustToken}) }}. TrustToken builds a security block to provide a root of trust-based IP security: secure key generation and truly random source. \textbf{TrustToken} only allows trusted communication between the non-trusted third-party IP and the rest of the SoC world by providing essential security features, i.e., secure, isolated execution, and trusted user interaction. The proposed design achieves this by interconnecting the third-party IP interface to \textbf{TrustToken} Controller and checking IP authorization(Token) signals \texttt{`correctness'} at run-time. \textbf{TrustToken} architecture shows a very low overhead resource utilization LUT (618, 1.16 \%), FF (44, 0.04 \%), and BUFG (2 , 6.25\%) in implementation. The experiment results show that TrustToken can provide a secure, low-cost, and trusted solution for non-trusted SoC IPs. △ Less

Submitted 26 September, 2022; originally announced September 2022.

Trusted IP solution in multi-tenant cloud FPGA platform

Authors: Muhammed Kawser Ahmed, Sujan Kumar Saha, Christophe Bobda

Abstract: Because FPGAs outperform traditional processing cores like CPUs and GPUs in terms of performance per watt and flexibility, they are being used more and more in cloud and data center applications. There are growing worries about the security risks posed by multi-tenant sharing as the demand for hardware acceleration increases and gradually gives way to FPGA multi-tenancy in the cloud. The confident… ▽ More Because FPGAs outperform traditional processing cores like CPUs and GPUs in terms of performance per watt and flexibility, they are being used more and more in cloud and data center applications. There are growing worries about the security risks posed by multi-tenant sharing as the demand for hardware acceleration increases and gradually gives way to FPGA multi-tenancy in the cloud. The confidentiality, integrity, and availability of FPGA-accelerated applications may be compromised if space-shared FPGAs are made available to many cloud tenants. We propose a root of trust-based trusted execution mechanism called \textbf{TrustToken} to prevent harmful software-level attackers from getting unauthorized access and jeopardizing security. With safe key creation and truly random sources, \textbf{TrustToken} creates a security block that serves as the foundation of trust-based IP security. By offering crucial security characteristics, such as secure, isolated execution and trusted user interaction, \textbf{TrustToken} only permits trustworthy connection between the non-trusted third-party IP and the rest of the SoC environment. The suggested approach does this by connecting the third-party IP interface to the \textbf{TrustToken} Controller and running run-time checks on the correctness of the IP authorization(Token) signals. With an emphasis on software-based assaults targeting unauthorized access and information leakage, we offer a noble hardware/software architecture for trusted execution in FPGA-accelerated clouds and data centers. △ Less

Submitted 22 September, 2022; originally announced September 2022.

Multi-Tenant Cloud FPGA: A Survey on Security

Authors: Muhammed Kawser Ahmed, Joel Mandebi, Sujan Kumar Saha, Christophe Bobda

Abstract: With the exponentially increasing demand for performance and scalability in cloud applications and systems, data center architectures evolved to integrate heterogeneous computing fabrics that leverage CPUs, GPUs, and FPGAs. FPGAs differ from traditional processing platforms such as CPUs and GPUs in that they are reconfigurable at run-time, providing increased and customized performance, flexibilit… ▽ More With the exponentially increasing demand for performance and scalability in cloud applications and systems, data center architectures evolved to integrate heterogeneous computing fabrics that leverage CPUs, GPUs, and FPGAs. FPGAs differ from traditional processing platforms such as CPUs and GPUs in that they are reconfigurable at run-time, providing increased and customized performance, flexibility, and acceleration. FPGAs can perform large-scale search optimization, acceleration, and signal processing tasks compared with power, latency, and processing speed. Many public cloud provider giants, including Amazon, Huawei, Microsoft, Alibaba, etc., have already started integrating FPGA-based cloud acceleration services. While FPGAs in cloud applications enable customized acceleration with low power consumption, it also incurs new security challenges that still need to be reviewed. Allowing cloud users to reconfigure the hardware design after deployment could open the backdoors for malicious attackers, potentially putting the cloud platform at risk. Considering security risks, public cloud providers still don't offer multi-tenant FPGA services. This paper analyzes the security concerns of multi-tenant cloud FPGAs, gives a thorough description of the security problems associated with them, and discusses upcoming future challenges in this field of study. △ Less

Submitted 22 September, 2022; originally announced September 2022.

Robust Artificial Delay based Impedance Control of Robotic Manipulators with Uncertain Dynamics

Authors: Udayan Banerjee, Bhabani Shankar Dey, Indra Narayan Kar, Subir Kumar Saha

Abstract: In this paper an artificial delay based impedance controller is proposed for robotic manipulators with uncertainty in dynamics. The control law unites the time delayed estimation (TDE) framework with a second order switching controller of super twisting algorithm (STA) type via a novel generalized filtered tracking error (GFTE). While time delayed estimation framework eliminates the need for accur… ▽ More In this paper an artificial delay based impedance controller is proposed for robotic manipulators with uncertainty in dynamics. The control law unites the time delayed estimation (TDE) framework with a second order switching controller of super twisting algorithm (STA) type via a novel generalized filtered tracking error (GFTE). While time delayed estimation framework eliminates the need for accurate modelling of robot dynamics by estimating the uncertain robot dynamics and interaction forces from immediate past data of state and control effort, the second order switching control law in the outer loop provides robustness against the time delayed estimation (TDE) error that arises due to approximation of the manipulator dynamics. Thus, the proposed control law tries to establish a desired impedance model between the robot end effector variables i.e. force and motion in presence of uncertainties, both when it is encountering smooth contact forces and during free motion. Simulation results for a two link manipulator using the proposed controller along with convergence analysis are shown to validate the proposition. △ Less

Submitted 20 August, 2022; v1 submitted 18 August, 2022; originally announced August 2022.

Fast Geometric Surface based Segmentation of Point Cloud from Lidar Data

Authors: Aritra Mukherjee, Sourya Dipta Das, Jasorsi Ghosh, Ananda S. Chowdhury, Sanjoy Kumar Saha

Abstract: Mapping the environment has been an important task for robot navigation and Simultaneous Localization And Mapping (SLAM). LIDAR provides a fast and accurate 3D point cloud map of the environment which helps in map building. However, processing millions of points in the point cloud becomes a computationally expensive task. In this paper, a methodology is presented to generate the segmented surfaces… ▽ More Mapping the environment has been an important task for robot navigation and Simultaneous Localization And Mapping (SLAM). LIDAR provides a fast and accurate 3D point cloud map of the environment which helps in map building. However, processing millions of points in the point cloud becomes a computationally expensive task. In this paper, a methodology is presented to generate the segmented surfaces in real time and these can be used in modeling the 3D objects. At first an algorithm is proposed for efficient map building from single shot data of spinning Lidar. It is based on fast meshing and sub-sampling. It exploits the physical design and the working principle of the spinning Lidar sensor. The generated mesh surfaces are then segmented by estimating the normal and considering their homogeneity. The segmented surfaces can be used as proposals for predicting geometrically accurate model of objects in the robots activity environment. The proposed methodology is compared with some popular point cloud segmentation methods to highlight the efficacy in terms of accuracy and speed. △ Less

Submitted 6 May, 2020; originally announced May 2020.

Comments: Accepted to PReMI 2019( Pattern Recognition and Machine Intelligence 2019). International Conference on Pattern Recognition and Machine Intelligence. Springer, Cham, 2019