A domain-equalized dual-stream CNN framework for robust synthetic dermatology detection, achieving 0.9971 AUC and 0.9277 OOD AUC.
Recommended citation: A. Kumar*, L. Chhetri*, D. Das*. (2026). "SPECTRAFORGE: Domain-Equalized Frequency-Spatial Fusion for Synthetic Dermatology Detection." IEEE DSAA.
An interpretable vision framework that automatically extracts human-readable fuzzy rules from deep visual representations for solar panel defect severity detection.
Recommended citation: L. Chhetri, A. Kumar, D. Das, P. Ghosal. (2026). "Interpretable Solar Panel Defect Detection via Fuzzy Rule Extraction from Hierarchical Vision Models." ICVGIP (Under Review).
A rigorous statistical approach applying Marginal Split Conformal Prediction to transformer-based satellite change detection, achieving distribution-free coverage bounds and SOTA F1 scores.
Recommended citation: A. Kumar, A. Mukherjee, S. Bala, S.R. Varma, A. Anand, H. Das, L. Chhetri. (2026). "Risk-Controlled Urban Change Detection: Conformal Prediction Wrappers for Provable Reliability in High-Resolution Satellite Imagery." ICCI.
An intrinsic defense architecture (DDL-FW) for Cyber-Physical Systems that embeds anomaly detection directly into residual connections, achieving an F1 score of 0.7206 on the HAI 21.03 benchmark.
Recommended citation: R. Das, L. Chhetri, A. Kumar, P. Ghosal. (2026). "Intrinsic Neural Firewalls for Cyber-Physical Systems: Robust Anomaly Rejection via Deep Delta Residual Overwrites." WIN 6.0.
We propose an Attention-Enhanced Swin Transformer integrating hierarchical windowed attention with CBAM to mitigate data leakage, achieving 96.82% accuracy on strict patient-level splits.
Recommended citation: L. Chhetri, P. Ghosal, A. Datta. (2026). "Attention-Enhanced Swin Transformers for Robust Brain Tumor Classification Under Patient-Level Data Splitting." IEEE GCON.
An uncertainty-guided semi-supervised learning framework that integrates Monte Carlo dropout-based epistemic uncertainty estimation to reduce silent failure rates by 29.2% in resource-constrained environments.
Recommended citation: A. Kumar, L. Chhetri, A. Anand, H. Das, A. Mukherjee. (2026). "Beyond Limited Labels: Safe Semi-Supervised Learning for Malaria Diagnosis." ICVGIP (Under Review).
A novel hierarchical gated convolutional architecture achieving real-time processing (853 FPS) and 95.9% accuracy for space manufacturing defect detection.
Recommended citation: L. Chhetri, A. Kumar. (2026). "DDV-GNet: High-Throughput Defect Detection for Space Manufacturing via Deep Delta Gated Networks." IEEE SPACE.
A comprehensive ablation study isolating the impact of specific data preprocessing and augmentation pipelines, proposing a novel “destructive vs. constructive” taxonomy for medical image processing.
Recommended citation: L. Chhetri, A. Kumar. (2026). "Optimizing Deep Learning for Brain Tumor Classification: A Comparative Ablation Study of Preprocessing and Augmentation Strategies." IEEE GCON.
A hierarchical architecture integrating 2D Deep Delta Learning, achieving 510 FPS and outperforming ResNet50 by 13.1x in deployment efficiency for satellite edge computing.
Recommended citation: L. Chhetri, A. Kumar, P. Ghosal (2026). "DeltaGate-CNN: A Lightweight Convolutional Architecture for Efficient Remote Sensing Image Classification." IEEE INDICON (Under Review).
A CNN-DDL architecture featuring a dynamic beta-gate and Marginal Split Conformal Prediction for real-time solar wind modeling, achieving zero-shot cross-solar-cycle generalization.
Recommended citation: A. Kumar, L. Chhetri, P. Ghosal. (2026). "Zero-Shot Cross-Solar Cycle Generalization of Ionospheric Total Electron Content Dynamics using a Deep Delta Learning Convolutional Neural Network." Advances in Space Research (Under Review).
An end-to-end OCR and visual inpainting framework for ancient Indic scripts, introducing a 150K sequence dataset generated via physically-motivated WGAN-GP degradation.
Recommended citation: L. Chhetri, et al. (2026). "Degraded Ancient Ashokan Brahmi Script Recognition via Self-Supervised Pretraining and WGAN-GP Degradation Pipelines." In Progress.
Integrating a learnable Radial Basis Function (RBF) fuzzy membership kernel into a hierarchical U-Net to dynamically filter semantic uncertainty, achieving SOTA IoU on skin lesion datasets.
Recommended citation: L. Chhetri, et al. (2026). "AHF-RBFNet: Attention-Guided Hierarchical Fusion U-Net with Learnable Fuzzy Membership Kernels for Medical Image Segmentation." In Progress.
An end-to-end AI analytics platform for property valuation using XGBoost and Streamlit.