Sankarasrinivasan Seshathiri

Sankarasrinivasan Seshathiri

Auto-generated profile, based on publicly available information

Sankarasrinivasan Seshathiri

Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei 106335, Taiwan

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2015 in Procedia Computer Science
Verified
Journal Article
Open access
SSS. Sankarasrinivasan
S. Sankarasrinivasan
BEBalasubramanian Esakki
Balasubramanian Esakki
KKK. Karthik
K. Karthik
UCU. Chandrasekar
U. Chandrasekar
Rishi GuptaRishi Gupta
Rishi Gupta
This paper describes an innovative protocol for full field mapping of a large civil structures involving effective use of Unmanned Aerial Vehicles (UAVs) to enable real time structural health monitoring. The proposed frameworks integrates UAVs, image processing and data acquisition procedures for crack detection and assessment of surface degradation. A novel approach is proposed combining hat transform and HSV thresholding technique for crack detection. In addition, grey scale thresholding is employed for the measurement of surface degradations. A Demonstration multi-rotor UAV model is developed to carry out full field inspection of civil structures and real time testing is performed in our large university campus. In order to provide sophisticated monitoring platform for users, a MATLAB Graphical User Interface (GUI) is developed to analyse real time as well as acquired images and the results are validated successfully. The obtained results confirms that, the envisaged approach is the foundation for cost effective and time compressing solution for monitoring of large civil structures

1 April 2018 in Journal of Performance of Constructed Facilities
Verified
Journal Article
Closed access
HPHaran Pragalath
Haran Pragalath
SSSankarasrinivasan Seshathiri
Sankarasrinivasan Seshathiri
HRHarsh Rathod
Harsh Rathod
BEBalasubramanian Esakki
Balasubramanian Esakki
Rishi GuptaRishi Gupta
Rishi Gupta
The safety and serviceability of civil infrastructures have to be ensured either as part of a periodic inspection program or immediately following a given hazardous event. The use of digital imaging techniques to identify the deformed or deteriorated surfaces of structures is a substantial area of research and aims to investigate a number of unknown parameters, including damage quantification and condition rating. This manuscript illustrates the integration of previously developed fuzzy logic–based decision-making tools with the currently developed image processing algorithm to quantify the damage for the condition rating of civil infrastructures. The proposed integrated framework exploits visual specifics of different elements of the infrastructure to perform automated evaluation of structural anomalies such as cracks and surface degradation. Two different image segmentation tools, (1) bottom hat transform and (2) hue, saturation, color (HSV) thresholding, are applied to identify the surface defects. The developed image processing software is used with the fuzzy set framework proposed in the previous research to gauge the damage indices due to various deterioration types like corrosion, alkali aggregate reaction, freeze–thaw attack, sulfate attack, acid attack or loading, fatigue, shrinkage, and honeycombing. Case studies of a long-span bridge and a warehouse building are illustrated for concept validation. The refined comprehensive method is presented as a graphical user interface (GUI) to facilitate the real-time condition assessment of civil infrastructures.

18 April 2023 in IEEE Transactions on Circuits and Systems for Video Technology
Verified
Journal Article
Closed access
Jing-Ming GuoJing-Ming Guo
Jing-Ming Guo
APAlim Wicaksono Hari Prayuda
Alim Wicaksono Hari Prayuda
Heri PrasetyoHeri Prasetyo
Heri Prasetyo
SSSankarasrinivasan Seshathiri
Sankarasrinivasan Seshathiri
Unsupervised image hashing is a widely used technique for large-scale image retrieval. This technique maps an image to a finite length of binary codes without extensive human-annotated data for compact storage and effective semantic retrieval. This study proposes a novel deep unsupervised double-bit hashing method for image retrieval. This approach is based on the double-bit hashing method, which has been shown to better preserve the neighboring structure of binary codes than single-bit hashing. Traditional double-bit hashing methods require the entire dataset to be processed simultaneously to determine optimal thresholding values of binary feature encoding. In contrast, the proposed method trains the hashing layer in a minibatch manner, allowing for adaptive threshold learning through a gradient-based optimization strategy. Additionally, unlike most former methods, which only train the hashing networks on top of fixed pre-trained neural networks backbone. The proposed learning framework trains both hashing and backbone networks alternately asynchronously. This strategy enables the model to maximize the learning capability of the hashing and backbone networks. Furthermore, adopting the lightweight Vision Transformer (ViT) in the proposed method allows the model to capture both local and global relationships between multiple image views exemplar, which lead to better generalization, thus maximizing the retrieval performance of the model. Extensive experiments on CIFAR10, NUW-WIDE, and FLICKR25K datasets validate that the proposed method has superior retrieval quality and computational efficiency than state-of-the-art methods.

14 February 2022 in Electronics
Verified
Journal Article
Open access
PDF file
Jing-Ming GuoJing-Ming Guo
Jing-Ming Guo
JYJr-Sheng Yang
Jr-Sheng Yang
SSSankarasrinivasan Seshathiri
Sankarasrinivasan Seshathiri
HWHung-Wei Wu
Hung-Wei Wu
This study details the development of a lightweight and high performance model, targeting real-time object detection. Several designed features were integrated into the proposed framework to accomplish a light weight, rapid execution, and optimal performance in object detection. Foremost, a sparse and lightweight structure was chosen as the network’s backbone, and feature fusion was performed using modified feature pyramid networks. Recent learning strategies in data augmentation, mixed precision training, and network sparsity were incorporated to substantially enhance the generalization for the lightweight model and boost the detection accuracy. Moreover, knowledge distillation was applied to tackle dropping issues, and a student–teacher learning mechanism was also integrated to ensure the best performance. The model was comprehensively tested using the MS-COCO 2017 dataset, and the experimental results clearly demonstrated that the proposed model could obtain a high detection performance in comparison to state-of-the-art methods, and required minimal computational resources, making it feasible for many real-time deployments.

1 November 2018
Verified
Conference Paper
Closed access
JGJing Ming Guo
Jing Ming Guo
SSS Sankarasrinivasan
S Sankarasrinivasan

26 April 2023 in Electronics
Verified
Journal Article
Open access
PDF file
Jing-Ming GuoJing-Ming Guo
Jing-Ming Guo
YHYu-Ting Hsiao
Yu-Ting Hsiao
Wei-Wen HsuWei-Wen Hsu
Wei-Wen Hsu
SSSankarasrinivasan Seshathiri
Sankarasrinivasan Seshathiri
Jiann-Der LeeJiann-Der Lee
Jiann-Der Lee
PLPeizhong Liu
Peizhong Liu
Earlier studies focused on training ResNet50 deep learning models on a dataset of fundus images from the National Taiwan University Hospital HsinChu Branch. The study aimed to identify class-specific discriminative areas related to various conditions of ganglion cell complex (GCC) thickness, center focus areas, cropped patches from the fundus, and dataset partitions. The study utilized two visualization methods to evaluate and explain the areas of interest of the network model and determine if they aligned with clinical diagnostic knowledge. The results of the experiments demonstrated that incorporating GCC thickness information improved the accuracy of glaucoma determination. The deep learning models primarily focused on the optic nerve head (ONH) for glaucoma diagnosis, which was consistent with clinical rules. Nonetheless, the models achieved high prediction accuracy in detecting glaucomatous cases using only cropped images of macular areas. Moreover, the model’s focus on regions with GCC impairment in some cases indicates that deep learning models can identify morphologically detailed alterations in fundus photographs that may be beyond the scope of visual diagnosis by experts. This highlights the significant contribution of deep learning models in the diagnosis of glaucoma.