System and method for identifying manufacturing defects
AssigneeSamsung Display Co., Ltd.
InventorsSai MarapaReddy, Shuhui Qu, Janghwan Lee
Life of this patent
Estimating the term…
Abstract
A system and method for classifying products manufactured via a manufacturing process. A processor receives an input dataset, and extracts features of the input dataset at two or more levels of abstraction. The processor combines the extracted features and provides the combined extracted features to a classifier. The classifier is trained based on the combined extracted features for learning a pattern of not-faulty products. The trained classifier is configured to receive data for a product to be classified, to output a prediction for the product based on the received data.
Claims20
A method for classifying products manufactured via a manufacturing process, the method comprising:
- receiving, by a processor, an input dataset;
- extracting, by the processor, features of the input dataset at two or more levels of abstraction, wherein the extracting of the features includes invoking a pre-trained neural network and extracting features from selected layers of the pre-trained neural network;
- combining, by the processor, the extracted features;
- providing, by the processor, the combined extracted features to a classifier;
- training, by the processor, the classifier based on the combined extracted features for learning a pattern of not-faulty products, the trained classifier being configured to receive data for a product to be classified, to output a prediction for the product based on the received data.
A system for classifying products manufactured via a manufacturing process, the system comprising:
- processor;
- and
- memory, wherein the memory has stored therein instructions that, when executed by the processor, cause the processor to:
- receive an input dataset;
- extract features of the input dataset at two or more levels of abstraction, wherein the instructions that cause the processor to extract the features further include instructions that cause the processor to invoke a pre-trained neural network and extract features from selected layers of the pre-trained neural network;
- combine the extracted features;
- provide the combined extracted features to a classifier;
- and
- train the classifier based on the combined extracted features for learning a pattern of only not-faulty products, the trained classifier being configured to receive data for a product to be classified, to output a prediction for the product based on the received data.
A non-transitory computer-readable medium storing instructions which, when executed by a processor, cause the processor to:
- receive an input dataset;
- extract features of the input dataset at two or more levels of abstraction, wherein the instructions that cause the processor to extract the features further include instructions that cause the processor to invoke a pre-trained neural network and extract features from selected layers of the pre-trained neural network;
- combine the extracted features;
- provide the combined extracted features to a classifier;
- train the classifier based on the combined extracted features for learning a pattern of not-faulty products, the trained classifier being configured to receive data for a product to be classified, to output a prediction for the product based on the received data.
Dependencies are inferred from the claim text.