Author

K. B

View on Semantic Scholar

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Open access Aug 2026

An Attention-Enhanced Lightweight CNN Framework with MTCNN Detection and TripletEmbedding Recognition for Occlusion-Robust Automated Attendance from Surveillance Video

Manual and card/barcode-based attendance recording remains slow, error-prone, and vulnerable to proxy marking, motivating fully automated, camera-based alternatives for schools and organizations. This paper proposes an Attention-Enhanced Lightweight CNN framework that couples MTCNN multi-scale face detection with a CBAM (Convolutional Block Attention Module) augmented MobileFace-style backbone trained under triplet loss to produce compact, discriminative 128-dimensional face embeddings from surveillance video. Enrolled identities are matched via cosine similarity against a reference embedding gallery, and a temporal multi-frame voting stage consolidates predictions across consecutive frames to suppress transient misdetections caused by pose change, partial occlusion, or motion blur.

K. B, L. C. · International Journal of Inn... · 0 citations

Conference Jul 2026

Interpretable Deep Learning Framework for Lung Cancer Classification using Attention and Visualization Techniques

Lung cancer has relatively higher prevalence amongst global deaths because of the delay in detecting cancerous tumors in patients. Detecting such tumors at an early stage using CT scan helps increase the chances of recovery of the patients. Interpretation of CT scans manually requires a significant amount of time and experience from the medical practitioners. In this study, we propose an explainable deep learning framework for automatically classifying multiple types of lung cancer through CT scans. The architecture of the network comprises of a DenseNet121 along with a Convolutional Block Attention Module (CBAM). Attention mechanisms were used on both channel and spatial aspects in succession to make sure that the model focuses on significant parts. A heatmap visualization approach is implemented through Grad-CAM to show the prediction made by the proposed model. These tests have been conducted on an open source dataset having a total of 4,598 CT scans of lungs categorized into Adenocarcinoma, Large Cell Carcinoma, and Normal Lung. The accuracy obtained by the suggested model is 94.6%, while the values of precision, recall, F1 score, and AUC are 0.95 and 0.96, respectively. The key reason behind the success of the suggested model over the other models such as VGG16, ResNet50, InceptionV3, and DenseNet121 is the utilization of the CBAM attention model.

Sivakarthi G, K. B, S. S et al. · International Conference Com... · 0 citations