Anirban Dasgupta
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Anirban Dasgupta has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
Aurobinda Routray
Indian Institute of Technology Kharagpur
7 shared publications- Aurobinda Routray
Department of Electrical Engineering, IIT Kharagpur, Kharagpur, India
5 shared publications Anjith George
2 shared publications- Anjith George
IDIAP Research Institute, Martigny, Switzerland
2 shared publications - Naragoni Saidulu
Department of Electronics Engineering, Indian Institute of Technology (BHU) Varanasi, India
2 shared publications - Priya Ranjan Muduli
Department of Electronics Engineering, Indian Institute of Technology (BHU) Varanasi, India
2 shared publications
Latest Publications
- 9 July 2026
- Published by Elsevier in Physica Medica
- Vol. 148, 105874
DARL-Net: Deformable-asymmetric residual learning and learnable non-local attention-based low-dose CT image denoising
- 9 July 2026
- Published by Elsevier in Physica Medica
- Vol. 148, 105874
DARL-Net: Deformable-asymmetric residual learning and learnable non-local attention-based low-dose CT image denoising
- 1 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Computer
- Vol. 59 (2) , 68-79
Exploring Recent Trends in Transfer Learning Strategies: A Case Study on Eye Localization and Eye State Classification
- 1 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Computer
- Vol. 59 (2) , 68-79
Exploring Recent Trends in Transfer Learning Strategies: A Case Study on Eye Localization and Eye State Classification
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 33 (2) , 1-11
LumiRoadGAN: Adaptive Generative Adversarial Network for Dark-to-Bright Road Image Translation
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 33 (2) , 1-11
LumiRoadGAN: Adaptive Generative Adversarial Network for Dark-to-Bright Road Image Translation
- 8 October 2025
- Published by Springer Nature in Journal of Global Entrepreneurship Research
- Vol. 15 (1) , 1-9
From doctoral research to entrepreneurship: a case study in bridging academia and entrepreneurship
- 8 October 2025
- Published by Springer Nature in Journal of Global Entrepreneurship Research
- Vol. 15 (1) , 1-9
From doctoral research to entrepreneurship: a case study in bridging academia and entrepreneurship
- 10 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Potentials
- Vol. 44 (2) , 24-31
Ocular diagnostics using electrooculography: Recent trends, challenges, and future directions
This article reviews the recent literature on the use of electrooculography (EOG) signals in diagnosing several ocular disorders. Since ocular disorders are becoming very common these days, it is essential to understand the role of EOG as a noninvasive technique to diagnose such conditions. This article introduces the principles of EOG signal acquisition, its typical characteristics, and the various paradigms of signal recording. The signal conditioning of EOG is discussed with respect to the latest techniques for removing noise and baseline drift. The use of the recent deep learning architectures for classifying EOG signals is discussed. The three major ocular conditions diagnosed with EOG are retinal, oculomotor, and vestibular dysfunctions. This article explores the manifestations of each condition in the EOG signal. This article further discusses subcategories of these conditions, such as the diagnosis of strabismus, nerve palsies, and vertigo-related disorders. A comparative analysis of EOG with other diagnostic methods, such as optical coherence tomography (OCT) and electroretinography (ERG), is provided. The review finally concludes by indicating the existing challenges.
- 10 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Potentials
- Vol. 44 (2) , 24-31
Ocular diagnostics using electrooculography: Recent trends, challenges, and future directions
This article reviews the recent literature on the use of electrooculography (EOG) signals in diagnosing several ocular disorders. Since ocular disorders are becoming very common these days, it is essential to understand the role of EOG as a noninvasive technique to diagnose such conditions. This article introduces the principles of EOG signal acquisition, its typical characteristics, and the various paradigms of signal recording. The signal conditioning of EOG is discussed with respect to the latest techniques for removing noise and baseline drift. The use of the recent deep learning architectures for classifying EOG signals is discussed. The three major ocular conditions diagnosed with EOG are retinal, oculomotor, and vestibular dysfunctions. This article explores the manifestations of each condition in the EOG signal. This article further discusses subcategories of these conditions, such as the diagnosis of strabismus, nerve palsies, and vertigo-related disorders. A comparative analysis of EOG with other diagnostic methods, such as optical coherence tomography (OCT) and electroretinography (ERG), is provided. The review finally concludes by indicating the existing challenges.
- 4 December 2024
- Published by Springer Nature
- p. 362-374
Optic Atrophy Classification from Fundus Images with Few-Shot Learning
- 4 December 2024
- Published by Springer Nature
- p. 362-374
Optic Atrophy Classification from Fundus Images with Few-Shot Learning
- 20 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-11
RHLNet: Robust Hybrid Loss-Based Network for Low-Dose CT Image Denoising
Low-dose computed tomography (LDCT) is a viable solution for clinical diagnosis despite noise and artifacts degrading diagnostic quality. Additionally, patients are protected from excessive X-ray radiation dose during the repeated CT scans obtained by normal-dose CT (NDCT). The data-driven deep learning algorithms with convolutional neural networks (CNN) are more effective at modeling structured CT data than random noise present. However, the aggressive denoising capability decimates structural characteristics. Also, the blurring effect may be introduced into reconstructed CT images via per-pixel-based objective functions such as mean squared error (MSE). Moreover, due to the perceptual loss computed on the feature space of the pre-trained VGG19 network, CT-specific details may be absent in perceptual features. In this study, we mitigate these problems with three novel accretions. First, we propose ensemble learning-based end-to-end low-dose CT denoising architecture using shallow and deep networks. We introduce novel lattice residual blocks (LRB) into a deep network architecture to retain long-range pixel dependency with weighted channel information. While preserving the local structure details, the three-layer shallow network accelerates the convergence of the deep network. Furthermore, we introduce Huber reconstruction loss, which combines the advantages of MSE and MAE for smooth and edge regions, respectively. We fine-tune VGG19 parameters to CT-specific data via transfer learning to avoid further loss of CT features. Finally, we include the Charbonnier loss as a structure loss term for improving visual quality. We perform extensive numerical experiments and analysis on the publicly available ‘’2016 NIH-AAPM-Mayo Clinic LDCT Grand Challenge” dataset.
- 20 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-11
RHLNet: Robust Hybrid Loss-Based Network for Low-Dose CT Image Denoising
Low-dose computed tomography (LDCT) is a viable solution for clinical diagnosis despite noise and artifacts degrading diagnostic quality. Additionally, patients are protected from excessive X-ray radiation dose during the repeated CT scans obtained by normal-dose CT (NDCT). The data-driven deep learning algorithms with convolutional neural networks (CNN) are more effective at modeling structured CT data than random noise present. However, the aggressive denoising capability decimates structural characteristics. Also, the blurring effect may be introduced into reconstructed CT images via per-pixel-based objective functions such as mean squared error (MSE). Moreover, due to the perceptual loss computed on the feature space of the pre-trained VGG19 network, CT-specific details may be absent in perceptual features. In this study, we mitigate these problems with three novel accretions. First, we propose ensemble learning-based end-to-end low-dose CT denoising architecture using shallow and deep networks. We introduce novel lattice residual blocks (LRB) into a deep network architecture to retain long-range pixel dependency with weighted channel information. While preserving the local structure details, the three-layer shallow network accelerates the convergence of the deep network. Furthermore, we introduce Huber reconstruction loss, which combines the advantages of MSE and MAE for smooth and edge regions, respectively. We fine-tune VGG19 parameters to CT-specific data via transfer learning to avoid further loss of CT features. Finally, we include the Charbonnier loss as a structure loss term for improving visual quality. We perform extensive numerical experiments and analysis on the publicly available ‘’2016 NIH-AAPM-Mayo Clinic LDCT Grand Challenge” dataset.
- 4 December 2023
- Published by Springer Nature
- p. 783-790
FVINet: Translating Thermograms to Facial Vasculature Imprints
- 4 December 2023
- Published by Springer Nature
- p. 783-790
FVINet: Translating Thermograms to Facial Vasculature Imprints
- Open Access
- 14 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. PP (15344320) , 1
A New Multi-resolution Analysis Method for Electrooculography Signals
Electrooculography (EOG) signals indicate the degree and direction of eye movements. Hence, EOG signals have been useful in eye movement controlled rehabilitation systems. Denoising and accurate identification of the type of eye movement in EOG signals are the major challenges in their analysis. The state-of-the-art techniques for EOG signal analysis concerning denoising and eye movement extraction are based on multi-resolution analysis using wavelet bases, such as Haar or Daubechies. However, these wavelets are designed for general purpose signal processing applications and hence are not optimized for the EOG signal structures. In this paper, we propose a new multi-resolution basis specific to the analysis of EOG signals. The scaling and wavelet functions for the basis are derived from the signatures of blinks and saccades respectively, and hence we name them as blinklets and saclets accordingly, thereby forming a new multi-resolution basis. These descriptors are found to be more effective than standard wavelets for EOG signals, signal denoising, and for identifying the different eye movement signatures such as saccades, blinks, smooth pursuits, and fixations, as tested on the Physiosig and Centre for Biomedical Cybernetics Eye Movement (CBC-EM) EOG Databases.
- 14 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. PP (15344320) , 1
Open AccessA New Multi-resolution Analysis Method for Electrooculography Signals
Electrooculography (EOG) signals indicate the degree and direction of eye movements. Hence, EOG signals have been useful in eye movement controlled rehabilitation systems. Denoising and accurate identification of the type of eye movement in EOG signals are the major challenges in their analysis. The state-of-the-art techniques for EOG signal analysis concerning denoising and eye movement extraction are based on multi-resolution analysis using wavelet bases, such as Haar or Daubechies. However, these wavelets are designed for general purpose signal processing applications and hence are not optimized for the EOG signal structures. In this paper, we propose a new multi-resolution basis specific to the analysis of EOG signals. The scaling and wavelet functions for the basis are derived from the signatures of blinks and saccades respectively, and hence we name them as blinklets and saclets accordingly, thereby forming a new multi-resolution basis. These descriptors are found to be more effective than standard wavelets for EOG signals, signal denoising, and for identifying the different eye movement signatures such as saccades, blinks, smooth pursuits, and fixations, as tested on the Physiosig and Centre for Biomedical Cybernetics Eye Movement (CBC-EM) EOG Databases.
- 1 November 2020
- Published by Elsevier in Pattern Recognition Letters
- Vol. 139, 104-111
Localization of eye Saccadic signatures in Electrooculograms using sparse representations with data driven dictionaries
- 1 November 2020
- Published by Elsevier in Pattern Recognition Letters
- Vol. 139, 104-111
Localization of eye Saccadic signatures in Electrooculograms using sparse representations with data driven dictionaries