Discover the Winners: Highlights from the Second Sony Spresense Student Developer Contest
Ph.D., Master, and Bachelor students from ETH Zürich and the University of Liverpool presented their projects on Tuesday, September 26, in the highly anticipated second edition of the Student Developer Contest. This competition was centered on the theme of "Tiny Machine Learning on Sony Spresense" and was designed to catalyze innovative concepts in the realm of smart wireless sensors.
The Winning Projects
The projects were evaluated by a panel of experts, comprising professors at the Universities and Sony engineers. Their expertise and discerning judgment ensured the competition's integrity and fairness. The three teams were awarded the following prizes: the first team, WH-1000XM5 headsets; the second SONY XE-300 linkbuds, and the third a SONY SRS-XE300 speaker.
From a pool of exceptional projects, three stood out for their innovation:
1st Place: Remote Health Monitoring System using Sony Spresense- This use case presents a project proposal to develop a wireless health monitoring system using the Sony Spresense microcontroller board. Its aim is to detect vital signs in real-time, offering benefits for remote health monitoring, especially during pandemics, and can aid healthcare, elderly care, and neonatal health diagnosis. They integrate various sensors with the Sony Spresense board and develop machine-learning algorithms for real-time vital sign detection.
2nd Place: Gaze Estimation on Spresense - This use case discusses the implementation of a gaze estimation system using the Sony Spresense microcontroller board. This was designed for applications in human-computer interaction, virtual reality, and medicine. The system employs a lightweight deep-learning model named TinyTrackerS, optimized for Spresense, running at 3 FPS. It showcases improved latency, energy efficiency, and model accuracy compared to previous models.
3rd Place: Tiny Object Detection on the Sony Spresense – This use case introduces a new technique for finding very small objects with the Sony Spresense microcontroller board it improves the Faster Objects More Objects (FOMO) network, making it better at spotting tiny objects. The method increases accuracy by 225% and reduces mistakes by 76% when tested on a data set of cars in parking lots. This is valuable for Internet of Things (IoT) applications that need to find small things efficiently on devices with limited resources.
The ETH Student Developer Contest is a contest organized by ETH Zurich University and the University of Liverpool, in collaboration with Sony’s Sensing Solutions Collaboration University Program (SSUP). SSUP is the university program launched by Sony Semiconductor Solutions Corporation (SSS) to build collaborations with university research groups around the world. It encourages and supports research projects with funding and hardware provided by Sony Semiconductor Solutions Corporation.
Congratulations to the winners and all who participated, and here´s to an exciting future that's full of opportunities in the field of technology and sensing solutions.
More information: