18595673. METHOD FOR PROVIDING A NEURAL NETWORK ON A DATA PROCESSING DEVICE simplified abstract (Robert Bosch GmbH)
Contents
METHOD FOR PROVIDING A NEURAL NETWORK ON A DATA PROCESSING DEVICE
Organization Name
Inventor(s)
Sebastian Boblest of Duernau (DE)
Benjamin Wagner of Friedrichshafen (DE)
Dennis Sebastian Rieber of Albstadt (DE)
Walid Hussien of Fellbach (DE)
METHOD FOR PROVIDING A NEURAL NETWORK ON A DATA PROCESSING DEVICE - A simplified explanation of the abstract
This abstract first appeared for US patent application 18595673 titled 'METHOD FOR PROVIDING A NEURAL NETWORK ON A DATA PROCESSING DEVICE
The abstract describes a method for providing a neural network on a data processing device by selecting an optimal implementation variant based on memory requirements and execution time.
- The method involves selecting a subset of implementation variants of the neural network that cannot be improved in terms of memory requirements, non-volatile memory requirement, and execution time without impairing the other two factors.
- Each subset contains at least one optimal implementation variant for each of the main memory requirement, non-volatile memory requirement, and execution time.
- The user can select one of the optimal implementation variants based on a user input specifying a selection from the subset.
- The selected implementation variant is then stored in the data processing device.
Potential Applications: - This method can be applied in various industries such as healthcare for medical image analysis, finance for fraud detection, and autonomous vehicles for object recognition. - It can also be used in consumer electronics for voice recognition and smart home devices for automation.
Problems Solved: - Addresses the challenge of optimizing neural network implementation variants for memory requirements and execution time without compromising performance. - Provides a user-friendly way to select the most suitable implementation variant for a specific application.
Benefits: - Improved efficiency in neural network deployment on data processing devices. - Enhanced performance without sacrificing memory requirements or execution time. - Simplified decision-making process for selecting optimal implementation variants.
Commercial Applications: Neural network optimization technology can be utilized in various commercial applications such as: - Edge computing devices for real-time processing. - Cloud computing services for enhanced performance. - IoT devices for efficient data processing.
Questions about Neural Network Optimization: 1. How does this method compare to traditional neural network deployment techniques? 2. What are the potential limitations of selecting an optimal implementation variant based on memory requirements and execution time?
Original Abstract Submitted
A method for providing a neural network on a data processing device. The method includes: ascertaining, from a set of implementation variants of the neural network, a subset with a plurality of implementation variants of the neural network, wherein each implementation variant of the subset cannot be improved with respect to any of main memory requirement, non-volatile memory requirement, and execution time, when executed on the data processing device, without impairing at least one of the other two, and the subset for each of main memory requirement, non-volatile memory requirement and execution time, when executed on the data processing device, contains at least one particular implementation variant that is optimal in this respect from the set of implementation variants; selecting one of the ascertained implementation variants according to a user input that specifies a selection from the subset; and storing the selected implementation variant in the data processing device.
