Mridula Verma

Mridula Verma

Auto-generated profile, based on publicly available information

Mridula Verma

Institute for Development and Research in Banking Technology, Hyderabad, India

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

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

Representative Publications

View All Publications
1 January 2021 in Computers and Electrical Engineering
Verified
Journal Article
Closed access
MMMahesh Kumar Morampudi
Mahesh Kumar Morampudi
MPMunaga V.N.K. Prasad
Munaga V.N.K. Prasad
MVMridula Verma
Mridula Verma
URU.S.N. Raju
U.S.N. Raju

7 August 2023 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
HSHemraj Singh
Hemraj Singh
MVMridula Verma
Mridula Verma
Ramalingaswamy CherukuRamalingaswamy Cheruku
Ramalingaswamy Cheruku
Appearance and motion are essential features in Video Salient Object Detection (VSOD) tasks. Most of the existing approaches utilize local features and thus fail to understand both the appearance as well as motion-specific semantics at the global level. Hence, these methods are unable to perform in unconstrained scenarios where multiple challenges, such as partial occlusion, motion blur, noise, clutter background, etc., exist. Moreover, these approaches require a large number of computational resources due to their complex structures, which limits their applicability to real-world deployment. To resolve these issues and to achieve a balance between accuracy and computational complexity, in this paper, a Dilation Separable Convolution Network (DSCNet) is proposed, which is equipped with Dilation Attention Fusion Module (DAFM), Bi-directional Cross-modality Fusion Module (BCFM), and Saliency Prediction Module (SPM) to extract enhanced multi-scaled motion and appearance features without increasing the model complexity. Further, a Bi-directional Separable Convolution Network (BSC-Net) equipped with a Separable Convolution Module (SCM)s and a FlowNet2.0 is proposed to utilize multi-scale contextual information across appearance cues and generate enhanced multi-scaled motion maps. For faster and better training of the DSCNet model, we propose a novel Stochastic Gradient-based Firefly Algorithm (SGFA), which adaptively balances the exploration and exploitation in multi-scaled, cross-modal embedded sub-spaces. With the help of the proposed SGFA algorithm, DSCNet+ model is constructed on top of DSCNet, which further improves the results in terms of the training speed as well as other evaluation metrics. The proposed models are evaluated on six benchmark datasets, and a detailed comparative study is provided with sixteen state-of-the-art (SOTA) models. One of the major highlights of the work is the significant performance of the proposed models on the most difficult DAVSOD-Diff dataset, which best reflects the challenging real-world scenarios.

30 September 2024 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
HSHemraj Singh
Hemraj Singh
MVMridula Verma
Mridula Verma
Ramalingaswamy CherukuRamalingaswamy Cheruku
Ramalingaswamy Cheruku
Geometric variations of spatial and temporal features of objects in video streams cause great difficulty in video salient object detection (VSOD) tasks. Most existing deep learning methods utilize fixed-sized kernels, which limits the receptive field to extract the local and global features and fails to understand the visual semantics of the deformed objects’ foreground and background. Moreover, due to their complex architectures, these methods need more computational resources, which limits their deployment in real-world scenarios. To address the aforementioned challenges and to make a balance between performance and computational complexity, a Deformable Separable Fusion Network (DSFNet) is proposed, which extracts the geometric spatiotemporal variations at multi-scale features dynamically without compromising the network’s complexity. A Swarm-Enhanced Adam (SEAdam) optimizer has been proposed to adaptively balance the exploration and exploitation of gradients locally and globally and improve the convergence speed. This is the first work that extracts the multi-scale geometric local and global context-based visual information. With the help of extensive experimentation on six benchmark highly challenging datasets, we show that the proposed model outperforms state-of-the-art models in terms of the number of parameters, floating point operations, and latency.

25 February 2026 in Data Mining and Knowledge Discovery
Verified
Journal Article
Closed access
PDF file
ZBZarka Bashir
Zarka Bashir
MVMridula Verma
Mridula Verma
CMC. Krishna Mohan
C. Krishna Mohan

30 April 2023
Verified
Conference Paper
Open access
HSHemraj Singh
Hemraj Singh
MVMridula Verma
Mridula Verma
Ramalingaswamy CherukuRamalingaswamy Cheruku
Ramalingaswamy Cheruku
The most challenging aspects of deploying deep models in IoT and embedded systems are extensive computational complexity and large training and inference time. Although various lightweight versions of state-of-the-art models are also being designed, maintaining the performance of such models is difficult. To overcome these problems, an efficient, lightweight, Deformable Separable Network (DSNet) is proposed for video salient object detection tasks, mainly for mobile and embedded vision applications. DSNet is equipped with a Deformable Convolution Network (DeCNet), Separable Convolution Network (SCNet), and Depth-wise Attention Response Propagation (DARP) module, which makes it maintain the trade-off between accuracy and latency. The proposed model generates saliency maps considering both the background and foreground simultaneously, making it perform better in unconstrained scenarios (such as partial occlusion, deformable background/objects, and illumination effect). The extensive experiments conducted on six benchmark datasets demonstrate that the proposed model outperforms state-of-art approaches in terms of computational complexity, number of parameters, and latency measures.

13 December 2017 in Applied Intelligence
Verified
Journal Article
Closed access
PDF file
MVMridula Verma
Mridula Verma
DSD. R. Sahu
D. R. Sahu
KSK. K. Shukla
K. K. Shukla