Mr. Abir Das | Artificial Intelligence | Research Excellence Award

Siliguri Government Polytechnic College | India

Abir Das is an emerging AI/ML researcher whose work spans deep learning, computer vision, medical imaging, and explainable AI. With a strong foundation in developing end-to-end AI systems, his research focuses on Vision Transformers, self-supervised learning, noisy-label correction, and interpretable models for high-stakes applications such as healthcare, EEG signal analysis, and industrial fault diagnosis. He has contributed as the first author to multiple international journals, working extensively on hybrid deep learning models, CLIP-based zero-shot learning, EEG motor imagery classification, and sensor-driven diagnostic pipelines. His research integrates expertise in PyTorch, TensorFlow, and modern transformer architectures, emphasizing human-centered, reliable, and transparent AI solutions. He has actively explored the intersection of computer vision and embedded systems, enhancing drone autonomy, depth estimation, and real-time object detection, while also contributing to speech technologies through accent-conversion and multimodal learning. His scientific output includes publications in reputable venues such as Scientific Reports, MDPI Sensors, and Computers, Materials & Continua. His growing scholarly impact is reflected in Scopus metrics: 11 citations from 11 documents with an h-index of 1, and Google Scholar metrics: 12 citations, h-index 1, i10-index 1. His work continues to advance practical and theoretically grounded AI methodologies, blending deep learning innovations with real-world applications across biomedical imaging, EEG analysis, and industrial AI systems.

Publication Profile

Scopus | Google Scholar

Featured Publications

Das, A., Singh, S., Kim, J., Ahanger, T. A., & Pisa, A. A. (2025). Enhanced EEG signal classification in brain computer interfaces using hybrid deep learning models. Scientific Reports, 15(1), 27161.

Zereen, A. N., Das, A., & Uddin, J. (2024). Machine fault diagnosis using audio sensor data and explainable AI techniques: LIME and SHAP. Computers, Materials & Continua, 80(3).

Das, S. S. A. (2025). Few-shot and zero-shot learning for MRI brain tumor classification using CLIP and Vision Transformers. Sensors, 25(23), 7341.