18633240. CALCULATION PROCESSING APPARATUS AND CALCULATION PROCESSING METHOD simplified abstract (CANON KABUSHIKI KAISHA)
Contents
CALCULATION PROCESSING APPARATUS AND CALCULATION PROCESSING METHOD
Organization Name
Inventor(s)
YUTAKA Murata of Kanagawa (JP)
MOTOKI Yoshinaga of Kanagawa (JP)
CALCULATION PROCESSING APPARATUS AND CALCULATION PROCESSING METHOD - A simplified explanation of the abstract
This abstract first appeared for US patent application 18633240 titled 'CALCULATION PROCESSING APPARATUS AND CALCULATION PROCESSING METHOD
The apparatus described in the patent application includes memories that hold feature planes corresponding to layers in a neural network, a calculation unit for processing these feature planes, and a memory control unit for managing the input and output of feature planes.
- Memories store feature planes for neural network layers
- Calculation unit processes feature planes
- Memory control unit reads and writes feature planes to memories
- Allows for connection of feature planes from different layers
- Facilitates seamless processing of feature planes in neural networks
Potential Applications: - Deep learning systems - Image and speech recognition technologies - Autonomous vehicles - Medical image analysis - Natural language processing
Problems Solved: - Efficient processing of feature planes in neural networks - Seamless integration of feature planes from different layers - Improved performance of deep learning algorithms
Benefits: - Enhanced accuracy in neural network processing - Faster computation of complex data - Facilitates advanced AI applications - Enables more sophisticated machine learning models
Commercial Applications: Title: "Advanced Neural Network Processing Apparatus" This technology can be utilized in industries such as: - Healthcare for medical imaging analysis - Automotive for autonomous driving systems - Finance for fraud detection and risk assessment - Retail for personalized recommendation systems
Questions about the technology: 1. How does this apparatus improve the efficiency of neural network processing? 2. What are the key advantages of connecting feature planes from different layers in a neural network?
Original Abstract Submitted
An apparatus includes memories hold feature planes each corresponding to a corresponding layer of layers in a neural network, a calculation unit performs calculation processing on the feature planes, and a memory control unit reads a feature plane from any of the memories and input the feature plane to the calculation unit, and writes a feature plane output from the calculation unit to any of the memories. In a case where feature planes corresponding to different layers are connected and the calculation processing is performed, the memory control unit writes the feature planes to be connected, in memories other than a specific memory among the memories, reads the feature planes to be connected, from the memories other than the specific memory and inputs the feature planes to the calculation unit, and writes the feature plane output from the calculation unit in the specific memory.
