RudraRoy
I am Rudra Roy, an AI engineer and researcher with interests spanning Computer Vision, Speech and Audio Processing, Medical Image Analysis, OCR, and multimodal deep learning. My research explores problems in sign-language recognition, speech enhancement, medical imaging, generative models, federated learning, and privacy-preserving AI, with a particular interest in translating methodological advances into usable real-world systems.
I currently work as a Machine Learning Engineer at ACI PLC, where I contribute to applied R&D in Bangla OCR, legal LLMs, Bangla ASR, document intelligence, and domain-focused AI systems.
Research
Exploring the frontiers of AI, signal processing, and embedded systems
Sign Language Recognition
Developing real-time sign language recognition systems using custom deep learning architectures and GAN-aided data augmentation for robust performance across diverse signing styles and environments.
Speech Enhancement
Building advanced speech enhancement models that operate in low-SNR conditions using novel architectures combining waveform processing with discrete wavelet transform for superior noise suppression.
IoT & Robotics
Designing embedded sensing systems and real-time monitoring solutions with control pipelines for practical industrial and agricultural applications, from greenhouse automation to gas burner safety.
Projects
Research projects and engineering builds
CAPRes50-GAN — Word-Level Sign Language Recognition
OngoingpFLOCT — Personalized FL for OCT Classification
OngoingCAR-UNet — Speech Enhancement with ConvNeXt + Attention
OngoingBreastDCGAN — End-to-End Breast Cancer Segmentation & Classification
OngoingPublications
Peer-reviewed research contributions
Enhancing Communication for the Deaf: Real-Time Sign Language Recognition & Translation
DPMAS-Net: Privacy-Preserving EMG Hand Gesture Recognition
