Patents by Inventor CHAO-MIN CHANG
CHAO-MIN CHANG has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 12315106Abstract: A booster engine enhances the quality of a frame sequence. The booster engine receives, from a first stage circuit, the frame sequence with quality degradation in at least a frame. The quality degradation includes at least one of uneven resolution and uneven frame per second (FPS). The booster engine queries an information repository for reference information on the frame, using a query input based on at least a region of the frame to obtain a query output. The booster engine then applies a neural network to the query input and the query output to generate an optimized frame, and sends an enhanced frame sequence including the optimized frame to a second stage circuit.Type: GrantFiled: September 7, 2022Date of Patent: May 27, 2025Assignee: MediaTek Inc.Inventors: Yao-Sheng Wang, Pei-Kuei Tsung, Chiani Lu, Chao-Min Chang, Yu-Sheng Lin, Wai Mun Wong
-
Patent number: 12272022Abstract: The quality of a frame sequence is enhanced by a booster engine collaborating with a first stage circuit. The first stage circuit adjusts the quality degradation of the frame sequence when a condition in constrained resources is detected. The quality degradation includes at least one of uneven resolution and uneven frame per second (FPS). The booster engine receives the frame sequence from the first stage circuit, and generates an enhanced frame sequence based on the frame sequence for transmission to a second stage circuit.Type: GrantFiled: August 24, 2022Date of Patent: April 8, 2025Assignee: MediaTek Inc.Inventors: Yao-Sheng Wang, Pei-Kuei Tsung, Chiung-Fu Chen, Wai Mun Wong, Chao-Min Chang, Yu-Sheng Lin, Chiani Lu, Chih-Cheng Chen
-
Patent number: 11983271Abstract: A processor may generate an enforcement point. The enforcement point may include one or more adversarial detection models. The processor may receive user input data. The processor may analyze, at the enforcement point, the user input data. The processor may determine, from the analyzing, whether there is an adversarial attack in the user input data. The processor may generate an alert based on the determining.Type: GrantFiled: November 19, 2020Date of Patent: May 14, 2024Assignee: International Business Machines CorporationInventors: Bruno dos Santos Silva, Cheng-Ta Lee, Ron Williams, Bo-Yu Kuo, Chao-Min Chang, Sridhar Muppidi
-
Patent number: 11768903Abstract: A computer-implemented method for automatically adjusting a Uniform Resource Locator (URL) seed list. The method includes crawling for documents based on a seed URL list. The method generates relations data from the documents using a Natural Language Processing (NLP) model. The method analyzes the relations data using an auto-seed model. The method modifies the seed URL list.Type: GrantFiled: June 19, 2020Date of Patent: September 26, 2023Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Chao-Min Chang, Ying-Chen Yu, June-Ray Lin, Kuei-Ching Lee, Curtis C H Wei
-
Patent number: 11663402Abstract: An approach for a fast and accurate word embedding model, “desc2vec,” for out-of-dictionary (OOD) words with a model learning from the dictionary descriptions of the word is disclosed. The approach includes determining that a target text element is not in a set of reference text elements, information describing the target text element is obtained. The information comprises a set of descriptive text elements. A set of vectorized representations for the set of descriptive text elements is determined. A target vectorized representation for the target text element is determined based on the set of vectorized representations using a machine learning model. The machine learning model is trained to represent a predetermined association between the set of vectorized representations for the set of descriptive text elements describing the target text element and the target vectorized representation.Type: GrantFiled: July 21, 2020Date of Patent: May 30, 2023Assignee: International Business Machines CorporationInventors: Chao-Min Chang, Kuei-Ching Lee, Ci-Hao Wu, Chia-Heng Lin
-
Publication number: 20230092969Abstract: An embodiment of the present invention is directed toward machine learning to produce results encompassing a new output. A machine learning model is trained to determine a candidate output from among a plurality of candidate outputs. First embeddings associated with the plurality of candidate outputs are generated from a first set of training data by an intermediate layer of the trained machine learning model. Second embeddings associated with a new candidate output are generated from a second set of training data by the intermediate layer of the trained machine learning model. A third embedding is determined for input data by the intermediate layer of the trained machine learning model. A resulting candidate output for the input data is predicted from a group of the plurality of candidate outputs and the new candidate output based on distances for the third embedding to the first and second embeddings.Type: ApplicationFiled: September 20, 2021Publication date: March 23, 2023Inventors: CHAO-MIN CHANG, Bo-Yu Kuo, Yu-Jin Chen, Yu-Chi Tang
-
Publication number: 20230087097Abstract: A booster engine enhances the quality of a frame sequence. The booster engine receives, from a first stage circuit, the frame sequence with quality degradation in at least a frame. The the quality degradation includes at least one of uneven resolution and uneven frame per second (FPS). The booster engine queries an information repository for reference information on the frame, using a query input based on at least a region of the frame to obtain a query output. The booster engine then applies a neural network to the query input and the query output to generate an optimized frame, and sends an enhanced frame sequence including the optimized frame to a second stage circuit.Type: ApplicationFiled: September 7, 2022Publication date: March 23, 2023Inventors: Yao-Sheng Wang, Pei-Kuei Tsung, Chiani Lu, Chao-Min Chang, Yu-Sheng Lin, Wai Mun Wong
-
Publication number: 20230067568Abstract: The quality of a frame sequence is enhanced by a booster engine collaborating with a first stage circuit. The first stage circuit adjusts the quality degradation of the frame sequence when a condition in constrained resources is detected. The quality degradation includes at least one of uneven resolution and uneven frame per second (FPS). The booster engine receives the frame sequence from the first stage circuit, and generates an enhanced frame sequence based on the frame sequence for transmission to a second stage circuit.Type: ApplicationFiled: August 24, 2022Publication date: March 2, 2023Inventors: Yao-Sheng Wang, Pei-Kuei Tsung, Chiung-Fu Chen, Wai Mun Wong, Chao-Min Chang, Yu-Sheng Lin, Chiani Lu, Chih-Cheng Chen
-
Publication number: 20230014551Abstract: A method for receiving a full training data set including a plurality of individual training data set, dividing the plurality of individual training sets into N classes, where N is an integer greater than three, dividing the N classes into M full data classes and N-M partial data classes, performing training to obtain a trained fixed size machine learning (ML) classification model and a trained in-class confidence model, outputting a first set of prediction value(s) based on the performance of training, distributing each class of the N classes of individual training data sets to a different node of a distributed machine learning system; and outputting, from the nodes of the distributed machine learning system, a second set of prediction value(s) for each class of the N classes.Type: ApplicationFiled: July 15, 2021Publication date: January 19, 2023Inventors: CHAO-MIN CHANG, Yu-Chi Tang, Bo-Yu Kuo, Yu-Jin Chen
-
Publication number: 20220230064Abstract: An analog circuit is calibrated to perform neural network computing. Calibration input is provided to a pre-trained neural network that includes at least a given layer having pre-trained weights stored in the analog circuit. The analog circuit performs tensor operations of the given layer using the pre-trained weights. Statistics of calibration output from the analog circuit is calculated. Normalization operations to be performed during neural network inference are determined. The normalization operations incorporate the statistics of the calibration output and are performed at a normalization layer that follows the given layer. A configuration of the normalization operations is written into memory while the pre-trained weights stay unchanged.Type: ApplicationFiled: January 6, 2022Publication date: July 21, 2022Inventors: Po-Heng Chen, Chia-Da Lee, Chao-Min Chang, Chih Chung Cheng, Hantao Huang, Pei-Kuei Tsung, Chun-Hao Wei, Ming Yu Chen
-
Publication number: 20220156376Abstract: A processor may generate an enforcement point. The enforcement point may include one or more adversarial detection models. The processor may receive user input data. The processor may analyze, at the enforcement point, the user input data. The processor may determine, from the analyzing, whether there is an adversarial attack in the user input data. The processor may generate an alert based on the determining.Type: ApplicationFiled: November 19, 2020Publication date: May 19, 2022Inventors: Bruno dos Santos Silva, Cheng-Ta Lee, Ron Williams, Bo-Yu Kuo, CHAO-MIN CHANG, Sridhar Muppidi
-
Publication number: 20220027557Abstract: An approach for a fast and accurate word embedding model, “desc2vec,” for out-of-dictionary (OOD) words with a model learning from the dictionary descriptions of the word is disclosed. The approach includes determining that a target text element is not in a set of reference text elements, information describing the target text element is obtained. The information comprises a set of descriptive text elements. A set of vectorized representations for the set of descriptive text elements is determined. A target vectorized representation for the target text element is determined based on the set of vectorized representations using a machine learning model. The machine learning model is trained to represent a predetermined association between the set of vectorized representations for the set of descriptive text elements describing the target text element and the target vectorized representation.Type: ApplicationFiled: July 21, 2020Publication date: January 27, 2022Inventors: Chao-Min Chang, Kuei-Ching Lee, Ci-Hao Wu, Chia-Heng Lin
-
Publication number: 20210397659Abstract: A computer-implemented method for automatically adjusting a Uniform Resource Locator (URL) seed list. The method includes crawling for documents based on a seed URL list. The method generates relations data from the documents using a Natural Language Processing (NLP) model. The method analyzes the relations data using an auto-seed model. The method modifies the seed URL list.Type: ApplicationFiled: June 19, 2020Publication date: December 23, 2021Inventors: CHAO-MIN CHANG, Ying-Chen Yu, June-Ray Lin, KUEI-CHING LEE, Curtis CH Wei