2025
1 Publication
IEEE Conference September 2025

Trans-REM: A Two Agent CNN-Transformer Based Approach for Indoor Radio Environment Mapping

S. Javid, S. Ghose, A. Dwivedi, and S. Sarkar

2025 36th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC 2025), Istanbul, Turkey

A novel hybrid CNN-Transformer framework for indoor radio environment mapping (REM). Trans-REM combines a CNN for local spatial feature extraction with a transformer for capturing global context. Two auxiliary input modalities — line-of-sight (LoS) image and antenna radiation pattern image — guide the network to model complex indoor propagation. Outperforms state-of-the-art REM methods by 15% in MSE.

2024
2 Publications
IEEE Conference November 2024

Efficiency Redefined: The Document Scanner App for Precise Entity Identification with Customized NER

S. Javid, M. Saim, and S. Islam

2024 3rd Edition of IEEE Delhi Section Flagship Conference (DELCON), New Delhi, India

Developed a specialized Named Entity Recognizer leveraging Computer Vision and NLP to extract entities from scanned documents. Achieved outstanding performance: Precision 99.80%, Recall 99.53%, F1 99.66%. Implemented as a privacy-first Document Scanner Web App combining efficient processing with robust data-privacy safeguards.

IEEE Conference September 2024

Enhancing Traffic Management Through Advanced Vehicle Detection for Congestion Prevention

M. Swaned, S. Javid, S. Humaney, A. Sachan, N. S. Chauhan, and N. Kumar

2024 IEEE 21st International Conference on Mobile Ad-Hoc and Smart Systems (MASS), Seoul, South Korea

Investigates urban traffic congestion and inefficiencies in traditional traffic management systems. Uses YOLOv8s deep learning for real-time detection of cars, two-wheelers, autos, buses, and trucks. Achieves 80% precision for cars and mAP@0.5 of 85.8%. Discusses integration with adaptive traffic light control to enhance traffic flow and urban safety.

2023
1 Publication
Journal Article August 2023

Spatio-Temporal Data Analysis using Deep Learning

A. Singh and S. Javid

International Research Journal of Engineering and Technology (IRJET), 10(8), 219–225

Explores the ability of deep learning to learn complex relationships between spatial and temporal dimensions. Evaluates deep learning methods applied to tasks in spatio-temporal analytics — transportation, social media event detection, environmental monitoring, human mobility prediction, action recognition, and related domains.