Patents by Inventor Hsi-Kang Tsao

Hsi-Kang Tsao 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).

  • Publication number: 20220129736
    Abstract: A mixed-precision quantization method for a neural network is provided. The neural network has a first precision and includes several layers and an original final output. For a particular layer, quantization of second precision on the particular layer and an input is performed. An output of the particular layer is obtained according to the particular layer of second precision and the input. De-quantization on the output of the particular layer is performed, and the de-quantized output is inputted to a next layer to obtain a final output. A value of an objective function is obtained according to the final output and the original final output. Above steps are repeated until the value of the objective function of each layer is obtained. A precision of quantization for each layer is decided according to the value of the objective function. The precision of quantization is one of first to fourth precision.
    Type: Application
    Filed: September 23, 2021
    Publication date: April 28, 2022
    Inventors: Bau-Cheng SHEN, Hsi-Kang TSAO, Chun-Yu LAI
  • Patent number: 9691381
    Abstract: An electronic device for browsing a document is disclosed. The document being browsed includes a plurality of command-associated text strings. First, a text string selector of the electronic device selects a plurality of candidate text strings from the command-associated text strings. Afterward, an acoustic string provider of the electronic device prepares a candidate acoustic string for each of the candidate text strings. Thereafter, a microphone of the electronic device receives a voice command. Next, a speech recognizer of the electronic device searches the candidate acoustic strings for a target acoustic string that matches the voice command, wherein the target acoustic string corresponds to a target text string of the candidate text strings. Finally, a document browser of the electronic device executes a command associated with the target text string.
    Type: Grant
    Filed: February 21, 2012
    Date of Patent: June 27, 2017
    Assignee: MEDIATEK INC.
    Inventors: Yiou-Wen Cheng, Liang-Che Sun, Chao-Ling Hsu, Hsi-Kang Tsao, Jyh-Horng Lin
  • Publication number: 20160050461
    Abstract: A multimedia processing apparatus includes a controller and a network module. The controller controls the multimedia processing apparatus to perform an application by setting control parameters to sub modules of the multimedia processing apparatus. The network module communicates to a social networking service server through a network. The controller is configured to generate and transmit at least one of setting information, favorite information and current viewing channel information to the social networking service server through the network module and the network. The setting information, the favorite information and current viewing channel information includes control parameters.
    Type: Application
    Filed: May 27, 2013
    Publication date: February 18, 2016
    Inventors: Xutong Hu, Hsi-Kang TSAO
  • Publication number: 20150016719
    Abstract: An embodiment of the invention provides a method performed by an electronic device. Under the method, the electronic device first analyzes significance-indication information of an image block of visual data to determine whether the image block encompasses a significant visual pattern. The significance-indication information includes the image block's time domain information and/or color domain information. Then, if the electronic device determines that the image block encompasses the significant visual pattern, it sifts the image block out from the visual data.
    Type: Application
    Filed: July 9, 2013
    Publication date: January 15, 2015
    Inventors: Io-Kuong Tam, Hsi-Kang Tsao
  • Publication number: 20130218573
    Abstract: An electronic device for browsing a document is disclosed. The document being browsed includes a plurality of command-associated text strings. First, a text string selector of the electronic device selects a plurality of candidate text strings from the command-associated text strings. Afterward, an acoustic string provider of the electronic device prepares a candidate acoustic string for each of the candidate text strings. Thereafter, a microphone of the electronic device receives a voice command. Next, a speech recognizer of the electronic device searches the candidate acoustic strings for a target acoustic string that matches the voice command, wherein the target acoustic string corresponds to a target text string of the candidate text strings. Finally, a document browser of the electronic device executes a command associated with the target text string.
    Type: Application
    Filed: February 21, 2012
    Publication date: August 22, 2013
    Inventors: Yiou-Wen Cheng, Liang-Che Sun, Chao-Ling Hsu, Hsi-Kang Tsao, Jyh-Horng Lin
  • Publication number: 20070171308
    Abstract: A method of color index searching in a color picture. An estimated color index X and an estimated range Y are obtained. An index range is set according to the estimated color index X and the estimated range Y. A color dimension of color signals corresponding to the index range is altered to make the color signals responding to the color dimension alteration. It is determined whether a designated region is responsive. If the designated region is not responsive, the estimated color index X is changed to update the index range for another color dimension alteration. If the designated region is responsive, the estimated range Y is reduced to update the index range for another color dimension alteration. If the estimated range reaches an acceptable range, the color index of the designated regions is found.
    Type: Application
    Filed: January 25, 2006
    Publication date: July 26, 2007
    Inventors: Chih-Lin Hsuan, Hong-Hui Chen, Hsi-Kang Tsao
  • Publication number: 20040179554
    Abstract: A method of implementing real-time video-audio interaction by data synchronization in an Internet game. The inventive method establishes an Internet transmission channel between Internet game clients without connecting to the Internet game sever. First, the method retrieves real-time video and audio data and compresses it into video and audio data packets. Next, the video and audio data packets are packaged into a transmission package. A time stamp is attached to the transmission package for data synchronization. The transmission package is transmitted to other game client through the Internet transmission channel. Finally, the Internet game client receives the transmission package, decompresses the received transmission package into video and audio data, and executes data synchronization according to the time stamp and output thereof.
    Type: Application
    Filed: March 11, 2004
    Publication date: September 16, 2004
    Inventor: Hsi-Kang Tsao
  • Publication number: 20040176168
    Abstract: A method of real-time video-audio interaction in an Internet game. An Internet transmission channel is established between Internet game clients without connection to the Internet game server. First, the method retrieves real-time video and audio data and compresses it into video and audio data packets. Next, the video and audio data packets are packaged into a transmission package to transmit to other game clients through the Internet transmission channel. Finally, the Internet game client receives the transmission package decompresses the received transmission package into video and audio data, and executes output thereof.
    Type: Application
    Filed: March 5, 2004
    Publication date: September 9, 2004
    Inventors: Hsi-Kang Tsao, Li-Hsiang Liao