Patents by Inventor Peizhao Li
Peizhao Li 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: 11886815Abstract: One example method involves operations for a processing device that include receiving, by a machine learning model trained to generate a search result, a search query for a text input. The machine learning model is trained by receiving pre-training data that includes multiple documents. Pre-training the machine learning model by generating, using an encoder, feature embeddings for each of the documents included in the pre-training data. The feature embeddings are generated by applying a masking function to visual and textual features in the documents. Training the machine learning model also includes generating, using the feature embeddings, output features for the documents by concatenating the feature embeddings and applying a non-linear mapping to the feature embeddings. Training the machine learning model further includes applying a linear classifier to the output features. Additionally, operations include generating, for display, a search result using the machine learning model based on the input.Type: GrantFiled: May 28, 2021Date of Patent: January 30, 2024Assignee: ADOBE INC.Inventors: Jiuxiang Gu, Vlad Morariu, Varun Manjunatha, Tong Sun, Rajiv Jain, Peizhao Li, Jason Kuen, Handong Zhao
-
Publication number: 20230305136Abstract: The present disclosure provides a system and a method for detecting and tracking objects. The method includes permuting an order of frames in a sequence of radar image frames to produce multiple permuted sequences with different frames at a dominant position in a corresponding permuted sequence of radar image frames. Each permuted sequence of radar image frames is processed with a first neural network to produce temporally enhanced features for each of the frames in the sequence of radar image frames. Further a feature map is reconstructed from the temporally enhanced features of each of the frames in the sequence of radar image frames to produce a sequence of feature maps. The method further includes processing a list of feature vectors from each feature map with a second neural network to produce temporally enhanced heatmaps.Type: ApplicationFiled: March 25, 2022Publication date: September 28, 2023Inventors: Pu Wang, Peizhao Li, Karl Berntorp
-
Publication number: 20220382975Abstract: One example method involves operations for a processing device that include receiving, by a machine learning model trained to generate a search result, a search query for a text input. The machine learning model is trained by receiving pre-training data that includes multiple documents. Pre-training the machine learning model by generating, using an encoder, feature embeddings for each of the documents included in the pre-training data. The feature embeddings are generated by applying a masking function to visual and textual features in the documents. Training the machine learning model also includes generating, using the feature embeddings, output features for the documents by concatenating the feature embeddings and applying a non-linear mapping to the feature embeddings. Training the machine learning model further includes applying a linear classifier to the output features. Additionally, operations include generating, for display, a search result using the machine learning model based on the input.Type: ApplicationFiled: May 28, 2021Publication date: December 1, 2022Inventors: Jiuxiang Gu, Vlad Morariu, Varun Manjunatha, Tong Sun, Rajiv Jain, Peizhao Li, Jason Kuen, Handong Zhao
-
Publication number: 20210174299Abstract: The present disclosure relates to the field of image processing technology, in particular to a method, system and device for collecting sales information of commodities in a cabinet, and a storage medium. Steps of the method include: successively acquiring pictures of the cabinet with the commodities placed therein in chronological order; acquiring commodity information on the former picture of every two adjacent cabinet pictures; successively adjusting every two adjacent cabinet pictures; comparing the corresponding commodity areas in every two adjacent cabinet pictures after the adjustment; and in the presence of different image contents, acquiring the number of commodity areas with different image contents and corresponding commodity information on the former picture.Type: ApplicationFiled: December 9, 2020Publication date: June 10, 2021Inventors: Xing Tong, Liang Sang, Yan Ke, Peizhao Li
-
Patent number: 8730298Abstract: A method for switching video resolution includes: confirming that a video sender and a video receiver have established a video communication connection, and obtaining an initial resolution of the video sender; confirming that bandwidth of the video communication connection and hardware of the video sender meet preset conditions, and switching the initial resolution of the video sender to a target resolution, in which the target resolution is higher than the initial resolution. After establishing video communication connections between terminals, when there are available bandwidth and hardware of the terminal using video communication, the resolution of terminal may be switched to provide a more clear video communication for terminal user, and guarantee smooth switching of resolution.Type: GrantFiled: July 12, 2012Date of Patent: May 20, 2014Assignee: Tencent Technology (Shenzhen) Company LimitedInventors: Chenchen Gu, Jing Lv, Xuecheng Luo, Peizhao Li, Pu Wang
-
Publication number: 20120281062Abstract: A method for switching video resolution includes: confirming that a video sender and a video receiver have established a video communication connection, and obtaining an initial resolution of the video sender; confirming that bandwidth of the video communication connection and hardware of the video sender meet preset conditions, and switching the initial resolution of the video sender to a target resolution, in which the target resolution is higher than the initial resolution. After establishing video communication connections between terminals, when there are available bandwidth and hardware of the terminal using video communication, the resolution of terminal may be switched to provide a more clear video communication for terminal user, and guarantee smooth switching of resolution.Type: ApplicationFiled: July 12, 2012Publication date: November 8, 2012Applicant: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Chenchen Gu, Jing Lv, Xuecheng Luo, Peizhao Li, Pu Wang
-
Publication number: 20110087730Abstract: Embodiments of this invention provide a method, device and system for creating an IM system window. Specifically, the method includes: pre-loading a dynamic link library and common resource file required for creating an IM system window; and when to create the IM system window for the first time currently, loading a characteristic resource file prepared for the IM system window, and creating the IM system window according to the pre-loaded dynamic link library and common resource file. In the embodiments of the present invention, before the user creates the IM system window for the first time, the dynamic link library and the common resource file required for creating the IM system window are pre-loaded. Hence, when the user creates the IM system window for the first time, the IM system window can be directly created according to the pre-loaded dynamic link library and common resource file.Type: ApplicationFiled: December 17, 2010Publication date: April 14, 2011Applicant: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Peizhao Li, Hong Jiang
-
Patent number: 7552181Abstract: The present invention is a method for implementing instant messaging, including: registering, in a Hook function, a function for registering an instant messaging sub-process in a current running program used for defining a condition for registering the instant messaging sub-process and a process for registering the instant messaging sub-process as a sub-process of the current running program if the condition is satisfied and is used for displaying a instant messaging sub-window in a window of the current running program; and if running an instant messaging program, calling, by the Hook function, the function for registering an instant messaging sub-process in a current running program, if the condition for registering the instant messaging sub-process is satisfied, creating an instant messaging sub-process, registering the sub-process as the sub-process of the current running program, and displaying the instant messaging sub-window in the current running program.Type: GrantFiled: January 19, 2005Date of Patent: June 23, 2009Assignee: Tencent Technology (Shenzhen) Company LimitedInventors: Qingming Wang, Yejun Huang, Songtao Lin, Peizhao Li
-
Publication number: 20070220091Abstract: The present invention is a method for implementing instant messaging, including: registering, in a Hook function, a function for registering an instant messaging sub-process in a current running program used for defining a condition for registering the instant messaging sub-process and a process for registering the instant messaging sub-process as a sub-process of the current running program if the condition is satisfied and is used for displaying a instant messaging sub-window in a window of the current running program; and if running an instant messaging program, calling, by the Hook function, the function for registering an instant messaging sub-process in a current running program, if the condition for registering the instant messaging sub-process is satisfied, creating an instant messaging sub-process, registering the sub-process as the sub-process of the current running program, and displaying the instant messaging sub-window in the current running program.Type: ApplicationFiled: January 19, 2005Publication date: September 20, 2007Inventors: Qingming Wang, Yejun Huang, Songtao Lin, Peizhao Li