Work

Code Optimization and Generation of Machine Learning and Driver Software for Memory-Constrained Edge Devices

Year: 2023
Type: dissertation
Abstract: This thesis contributes methods for optimization and generation of software for compact electronic edge devices, cruicial across diverse applications. Emphasizing memory constraints as pivotal factors influencing cost and power consumption, the methods presented within this thesis enhance machine learning applications, both on individual devices and within distributed scenarios. Moreover, these methods also improve driver software with memory reductions and automation.
Language: English
Open Access status: green
OpenAlex ID: W6980596664
Cites: 86
Cited by:
Related to: 10