18238016. CONVOLUTIONAL NEURAL NETWORK PRUNING PROCESSING METHOD, DATA PROCESSING METHOD, AND DEVICE simplified abstract (HUAWEI TECHNOLOGIES CO., LTD.)
Contents
CONVOLUTIONAL NEURAL NETWORK PRUNING PROCESSING METHOD, DATA PROCESSING METHOD, AND DEVICE
Organization Name
Inventor(s)
CONVOLUTIONAL NEURAL NETWORK PRUNING PROCESSING METHOD, DATA PROCESSING METHOD, AND DEVICE - A simplified explanation of the abstract
This abstract first appeared for US patent application 18238016 titled 'CONVOLUTIONAL NEURAL NETWORK PRUNING PROCESSING METHOD, DATA PROCESSING METHOD, AND DEVICE
Simplified Explanation
The abstract describes a method for pruning convolutional neural networks (CNNs) using a constructed objective loss function. The method involves performing sparse training on the CNN by using three sub-loss functions within the objective loss function.
- The method is used for pruning convolutional neural networks.
- It involves performing sparse training on the CNN.
- A constructed objective loss function is used, which includes three sub-loss functions.
- The objective loss function helps in identifying and removing unnecessary connections in the CNN.
- The method can be applied in the field of artificial intelligence.
Potential Applications
This technology has potential applications in various fields, including:
- Artificial intelligence research and development
- Computer vision tasks, such as image classification and object detection
- Deep learning applications, such as natural language processing and speech recognition
- Optimization of neural networks for efficient inference on resource-constrained devices
Problems Solved
The technology addresses the following problems:
- Reducing the complexity and computational requirements of convolutional neural networks
- Improving the efficiency and speed of neural network training and inference
- Enabling the deployment of neural networks on resource-limited devices with limited memory and processing power
- Enhancing the interpretability and explainability of neural networks by removing unnecessary connections
Benefits
The technology offers several benefits:
- Improved efficiency and speed of training and inference in convolutional neural networks
- Reduced memory and computational requirements, enabling deployment on resource-constrained devices
- Enhanced interpretability and explainability of neural networks by removing unnecessary connections
- Potential for improved accuracy and performance of neural networks through optimized pruning techniques
Original Abstract Submitted
Embodiments of this application disclose a convolutional neural network pruning processing method, a data processing method, and a device, which may be applied to the field of artificial intelligence. The convolutional neural network pruning processing method includes: performing sparse training on a convolutional neural network by using a constructed objective loss function, where the objective loss function may include three sub-loss functions.
