17957689. CNN SEAMLESS TILE PROCESSING FOR LOW-POWER INFERENCE ACCELERATOR simplified abstract (ADVANCED MICRO DEVICES, INC.)
Contents
- 1 CNN SEAMLESS TILE PROCESSING FOR LOW-POWER INFERENCE ACCELERATOR
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 CNN SEAMLESS TILE PROCESSING FOR LOW-POWER INFERENCE ACCELERATOR - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Original Abstract Submitted
CNN SEAMLESS TILE PROCESSING FOR LOW-POWER INFERENCE ACCELERATOR
Organization Name
Inventor(s)
Tung Chuen Kwong of Markham (CA)
Akila Subramaniam of Austin TX (US)
CNN SEAMLESS TILE PROCESSING FOR LOW-POWER INFERENCE ACCELERATOR - A simplified explanation of the abstract
This abstract first appeared for US patent application 17957689 titled 'CNN SEAMLESS TILE PROCESSING FOR LOW-POWER INFERENCE ACCELERATOR
Simplified Explanation
The patent application describes methods and devices for processing image data using layers of a convolutional neural network on a sub-frame portion basis. The processing device includes memory and a processor that determines the size of input tiles based on receptive fields and allocated local memory.
- The processor determines receptive fields for input tiles via backward propagation.
- The size of input tiles is determined based on receptive fields and allocated local memory.
- The processor checks if the allocated local memory is sufficient to store data for input tiles and padded data for receptive fields.
Potential Applications
This technology can be applied in image recognition systems, autonomous vehicles, medical imaging, and video processing.
Problems Solved
This technology solves the problem of efficiently processing image data in neural networks by optimizing memory usage for sub-frame portions.
Benefits
The benefits of this technology include improved efficiency in image data processing, reduced memory usage, and enhanced performance of convolutional neural networks.
Potential Commercial Applications
Potential commercial applications of this technology include in the fields of computer vision, artificial intelligence, robotics, and surveillance systems.
Possible Prior Art
One possible prior art could be the use of memory optimization techniques in neural networks for image processing tasks.
Unanswered Questions
How does this technology compare to existing methods for processing image data in neural networks?
This article does not provide a direct comparison with existing methods for processing image data in neural networks. Further research or a comparative study would be needed to address this question.
What are the specific limitations or constraints of this technology in practical applications?
The article does not mention any specific limitations or constraints of this technology in practical applications. Additional testing and real-world implementation may be necessary to identify and address any such limitations.
Original Abstract Submitted
Methods and devices are provided for processing image data on a sub-frame portion basis using layers of a convolutional neural network. The processing device comprises memory and a processor. The processor is configured to determine, for an input tile of an image, a receptive field via backward propagation and determine a size of the input tile based on the receptive field and an amount of local memory allocated to store data for the input tile. The processor determines whether the amount of local memory allocated to store the data of the input tile and padded data for the receptive field.
