Reza Rawassizadeh

Timeline

1. 2013

   Education and UbiqLog

   He holds a bachelor's degree from Tehran Azad University (Central Branch), an M.S. from Vienna University of Technology, and a Ph.D. in computer science from the University of Vienna. In 2013, he coauthors UbiqLog, a generic mobile phone-based life-log framework, with Martin Tomitsch, Katarzyna Wac and A Min Tjoa.

2. 2015

   Smartwatch context sensing

   In 2015, he coauthors “Energy-Efficient Integration of Continuous Context Sensing and Prediction into Smartwatches,” a framework that combines continuous context sensing with on-device prediction. The team evaluates energy use and prediction accuracy on smartwatch hardware.

3. 2017

   Speaker recognition on wearables

   In 2017, he coauthors “Toward Accurate and Efficient Feature Selection for Speaker Recognition on Wearables” with Rui Liu and David Kotz of Dartmouth College, for the Proceedings of the ACM Workshop on Wearable Systems and Applications. The paper compares four feature-selection algorithms and three feature sets for speaker identification and verification on small wearable platforms.

4. March 2018

   Vocal resonance authentication

   In March 2018, he coauthors “Vocal Resonance: Using Internal Body Voice for Wearable Authentication” with Rui Liu, Cory Cornelius, Ronald Peterson and David Kotz. A throat-mounted microphone supplies voice samples to an LSTM-based deep-learning model that reaches balanced accuracy of 0.914 for identification and 0.961 for verification across 29 subjects.

5. 2022

   Smartwatch heart-rate patterns

   In 2022, he coauthors a longitudinal study of heart rate and physical activity data from smartwatches worn by 1,014 users, with Fatemeh Karimi, Zohre Amoozgar, Reza Reiazi and Mehdi Hosseinzadeh. The analysis finds three common heart-rate patterns during sleep and two during the day.

6. 2022

   ODSearch wins a BCS award

   In 2022, he and Yi Rong win the BCS Search Industry Award for Best Search User Experience for ODSearch. The system searches large datasets on resource-constrained devices such as smartwatches in near real time.

7. 2025

   A textbook, GradES and TinyMusician

   In March 2025, he publishes Machine Learning and Artificial Intelligence: Concepts, Algorithms and Models, a 1,166-page book developed over eight years. Its chapters cover clustering, regression, classification, neural networks, self-supervised deep learning, reinforcement learning, and lighter neural-network and machine-learning models. His 2025 preprints report that GradES speeds transformer training by 1.57 to 7.22 times, while TinyMusician, distilled from MusicGen, retains 93% of MusicGen-Small performance with 55% less model size.

8. As of October 2026

   Directing the Resource-Efficient AI Lab

   In August 2026, he coauthors “Gradient Under Microscope: Benchmarking Resource Utilization of Memory-Efficient Gradient Computation Methods,” which compares five optimizers and three memory strategies across four transformer architectures. As of October 2026, he is an Associate Professor of Computer Science and Coordinator of AI and Machine Learning at Boston University's Metropolitan College, directs the Resource-Efficient AI Lab, and designs low-power machine-learning algorithms for small battery-powered devices such as smartwatches and fitness trackers.