Sayantan Dey is working as a Project Associate at Parimal Lab, Department of Computer Science and Engineering, IIT Roorkee, under the supervision of Dr. Partha Pratim Roy (IITR),  focusing on Natural Language Processing. His interests span various domains, including data analytics, machine learning, deep learning, computer vision, signal processing, and brain-computer interfaces.

Prior to his role at IIT Roorkee, Sayantan gained experience as a Programmer Analyst Trainee (Google Cloud Engineer) at Cognizant, followed by a tenure as a Machine Learning Engineer at the National Informatics Center. He holds a Bachelor of Technology in Electrical Engineering from St. Thomas’ College of Engineering & Technology, Kolkata.

Membership of Professional Societies

• IEI (The Institution of Engineers) Student Member. (Since 2020).

Projects

• Handwritten Text Recognition: It aims to develop an advanced and efficient system capable of accurately interpreting and digitizing handwritten text in various Indian languages. Given the diversity and complexity of scripts such as Devanagari, Bengali, Tamil, and many others, this project focuses on creating a robust recognition model using state-of-the-art deep learning techniques. It will address the unique challenges these scripts pose, including variations in writing styles, intricate ligatures, and contextual dependencies. This project contributes to computer vision and natural language processing and plays a pivotal role in enhancing digital literacy and information retrieval in a multilingual nation.

• Multilingual Text Recognition and Interpretation: Funded by the National Informatic Center (NIC): Multilingual text classification in Indian languages involves categorizing text documents into predefined categories based on their content. Given the linguistic diversity in India, this task requires handling multiple scripts and dialects. Advanced techniques like natural language processing (NLP) and machine learning are used to train models on large datasets encompassing various languages. These models can identify patterns and features specific to each language, improving the accuracy of text classification. Multilingual embeddings and transfer learning further enhance these models by allowing knowledge sharing across languages, leading to more robust and scalable text classification systems in the Indian multilingual context.

Publications

• Gourav Siddhad, Sayantan Dey, Partha Pratim Roy, Masakazu Iwamura. "Awake at the Wheel: Enhancing Automotive Safety through EEG-Based Fatigue Detection". 2024 27th International Conference on Pattern Recognition (ICPR), 2024.
  [Core2023: B]

• Sayantan Dey, Shivam Thakur, Akhilesh Kandwal, Rohit Kumar, Sharmistha Dasgupta, Partha Pratim Roy. "BharatBhasaNet-A unified framework to identify Indian code mix Languages." IEEE Access (2024).

• Sayanjit Singha Roy, Soumya Chatterjee, Ritwick Barman, Saptarshi Roy, Sayantan Dey. "Bearing fault detection in induction motors employing difference visibility graph." 2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES). IEEE, 2020.

• Sayanjit Singha Roy, Sayantan Dey, Soumya Chatterjee. "Autocorrelation aided random forest classifier-based bearing fault detection framework." IEEE Sensors Journal 20.18 (2020): 10792-10800.

• Sudip Modak, Sayanjit Singha Roy, Kaniska Samanta, Soumya Chatterjee, Sayantan Dey, Ronjoy Bhowmik, Rohit Bose. "Detection of focal EEG signals employing weighted visibility graph." 2020 International Conference on Computer, Electrical & Communication Engineering (ICCECE). IEEE, 2020.

• Sayantan Dey, Sayanjit Singha Roy, Kaniska Samanta, Sudip Modak, Soumya Chatterjee. "Autocorrelation based feature extraction for bearing fault detection in induction motors." 2019 International Conference on Electrical, Electronics and Computer Engineering (UPCON). IEEE, 2019.