Hui Jui Hu, 598325 Vietor Ave, Flushing, NY 11373

2021037, Dec 2, 2021

Jun 17, 2021

17/350917

- Kansas City MO, US
Hui Peng Hu - Oakland CA, US

G06F 40/30
G06N 3/08
G06F 40/263

A technique making use of a few-shot model to determine whether a query text content belongs to a same language as a small set of examples, or alternatively provide a next member in the same language to the small set of examples. The related few-shot model makes use of convolutional models that are trained in a “learning-to-learn” fashion such that the models know how to evaluate few-shots that belong to the same language. The term “language” in this usage is broader than spoken languages (e.g., English, Spanish, German, etc.). “Language” refers to a category, or data domain, of expression through characters. Belonging to a given language is not specifically based on what the language is, but the customs or traits expressed in that language.

2021004, Feb 18, 2021

Aug 2, 2019

16/530778

- Oakland CA, US
Yoriyasu Yano - Oakland CA, US
Hui Peng Hu - Berkeley CA, US
Kuang Chen - Oakland CA, US

G06K 9/46
G06F 16/583
G06K 9/00

Methods, apparatuses, and embodiments related to analyzing the content of digital images. A computer extracts multiple sets of visual features, which can be keypoints, based on an image of a selected object. Each of the multiple sets of visual features is extracted by a different visual feature extractor. The computer further extracts a visual word count vector based on the image of the selected object. An image query is executed based on the extracted visual features and the extracted visual word count vector to identify one or more candidate template objects of which the selected object may be an instance. When multiple candidate template objects are identified, a matching algorithm compares the selected object with the candidate template objects to determine a particular candidate template of which the selected object is an instance.

2020036, Nov 19, 2020

May 15, 2019

16/413159

- Oakland CA, US
Hui Peng Hu - Berkeley CA, US

G06F 17/27
G06N 3/08

A technique making use of a few-shot model to determine whether a query text content belongs to a same language as a small set of examples, or alternatively provide a next member in the same language to the small set of examples. The related few-shot model makes use of convolutional models that are trained in a “learning-to-learn” fashion such that the models know how to evaluate few-shots that belong to the same language. The term “language” in this usage is broader than spoken languages (e.g., English, Spanish, German, etc.). “Language” refers to a category, or data domain, of expression through characters. Belonging to a given language is not specifically based on what the language is, but the customs or traits expressed in that language.

2019022, Jul 18, 2019

Jan 16, 2019

16/249561

- Oakland CA, US
Hui Peng Hu - Berkeley CA, US

G06F 17/24
G06F 3/0481
G06F 16/583

Methods, apparatuses, and embodiments related to interactively predicting fields in a form. A computer system received an image of a form. A user moves a cursor to a first field of the form, and the computer system automatically displays a predicted location of the field, including a bounding box that represents a boundary of the field. The computer system further predicts the field name/label based on text in the document. The user clicks on the field to indicate that he wants to digitize the field. When needed, the user interactively modifies the size of the bounding box that represents the extent of the field, changes the name/label of the field. Once finalized, the user can cause the field information (e.g., the bounding box coordinate, the bounding box location, the name/label of the field, etc.) to be written to a database.

2019019, Jun 27, 2019

Dec 21, 2018

16/230812

- Oakland CA, US
Michail Iliadis - Oakland CA, US
Hui Peng Hu - Berkeley CA, US

G06K 9/00
G06K 9/62

Disclosed are a method and apparatus for determining a given variation of a form used by a filled in instance of that type of form from amongst a number of form templates. The given instance is aligned to each of the variants or form templates. The result of the alignment includes a series of key points that did not match up well (“bad” key points). The bad key points are taken from the form templates. Then, a set of pixel patches from around each of the bad key points of the form templates are extracted. The pixel patches are individually compared to corresponding pixel patches of the instance. The comparison generates a match score. The form template having the greatest match score is the correct form template.

2018014, May 24, 2018

Jan 19, 2018

15/875969

- Oakland CA, US
Hui Peng Hu - Berkeley CA, US

G06F 17/24
G06F 3/0481
G06F 17/30

Methods, apparatuses, and embodiments related to interactively predicting fields in a form. A computer system received an image of a form. A user moves a cursor to a first field of the form, and the computer system automatically displays a predicted location of the field, including a bounding box that represents a boundary of the field. The computer system further predicts the field name/label based on text in the document. The user clicks on the field to indicate that he wants to digitize the field. When needed, the user interactively modifies the size of the bounding box that represents the extent of the field, changes the name/label of the field. Once finalized, the user can cause the field information (e.g., the bounding box coordinate, the bounding box location, the name/label of the field, etc.) to be written to a database.

2017025, Sep 7, 2017

Apr 10, 2017

15/483291

- Oakland CA, US
Yoriyasu Yano - Oakland CA, US
Hui Peng Hu - Berkeley CA, US
Kuang Chen - Oakland CA, US

G06K 9/46
G06F 17/30

Methods, apparatuses, and embodiments related to analyzing the content of digital images. A computer extracts multiple sets of visual features, which can be keypoints, based on an image of a selected object. Each of the multiple sets of visual features is extracted by a different visual feature extractor. The computer further extracts a visual word count vector based on the image of the selected object. An image query is executed based on the extracted visual features and the extracted visual word count vector to identify one or more candidate template objects of which the selected object may be an instance. When multiple candidate template objects are identified, a matching algorithm compares the selected object with the candidate template objects to determine a particular candidate template of which the selected object is an instance.

2017004, Feb 16, 2017

Aug 12, 2015

14/824493

- Oakland CA, US
Hui Peng Hu - Berkeley CA, US

G06F 17/24
G06F 3/0481
G06F 17/30

Methods, apparatuses, and embodiments related to interactively predicting fields in a form. A computer system received an image of a form. A user moves a cursor to a first field of the form, and the computer system automatically displays a predicted location of the field, including a bounding box that represents a boundary of the field. The computer system further predicts the field name/label based on text in the document. The user clicks on the field to indicate that he wants to digitize the field. When needed, the user interactively modifies the size of the bounding box that represents the extent of the field, changes the name/label of the field. Once finalized, the user can cause the field information (e.g., the bounding box coordinate, the bounding box location, the name/label of the field, etc.) to be written to a database.