Daniel Dinu

US Patent:

20090002379, Jan 1, 2009

Filed:

Jun 30, 2007

Appl. No.:

11/824508

Inventors:

Juan Carlos Baeza - Bellevue WA, US
Eric S. Christoffersen - Beaux Arts WA, US
Sean M. Callahan - St. Paul MN, US
Daniel Dinu - Redmond WA, US
Barry Friemel - Redmond WA, US
William Chen - Issaquah WA, US
Weidong Zhao - Bellevue WA, US
Yongjun Wu - Seattle WA, US

Assignee:

Microsoft Corporation - Redmond WA

International Classification:

G06T 1/20

US Classification:

345522, 345501

Abstract:

Video decoding innovations for multithreading implementations and graphics processor unit (“GPU”) implementations are described. For example, for multithreaded decoding, a decoder uses innovations in the areas of layered data structures, picture extent discovery, a picture command queue, and/or task scheduling for multithreading. Or, for a GPU implementation, a decoder uses innovations in the areas of inverse transforms, inverse quantization, fractional interpolation, intra prediction using waves, loop filtering using waves, memory usage and/or performance-adaptive loop filtering. Innovations are also described in the areas of error handling and recovery, determination of neighbor availability for operations such as context modeling and intra prediction, CABAC decoding, computation of collocated information for direct mode macroblocks in B slices, reduction of memory consumption, implementation of trick play modes, and picture dropping for quality adjustment.

US Patent:

20200402182, Dec 24, 2020

Filed:

May 16, 2018

Appl. No.:

15/981848

Inventors:

- Menlo Park CA, US
Daniel Dinu - Sunnyvale CA, US

International Classification:

G06Q 50/00
G06F 11/34
G06F 17/30

Abstract:

Systems, methods, and non-transitory computer readable media are configured to determine a skill level of a user. One or more features can be selected based on the skill level. Subsequently, one or more of the features can be activated for a page on a social networking system.

US Patent:

20190213282, Jul 11, 2019

Filed:

Jan 11, 2018

Appl. No.:

15/868930

Inventors:

- Menlo Park CA, US
Daniel Dinu - Sunnyvale CA, US

International Classification:

G06F 17/30

Abstract:

Systems, methods, and non-transitory computer readable media are configured to receive a specification of an entity having a presence via an online channel. One or more scores based on one or more occurrences relating to the presence of the entity can be generated. The occurrences can relate to at least one of impressions or engagements by users in relation to the presence of the entity. Subsequently, one or more of the scores can be presented.

US Patent:

20190208025, Jul 4, 2019

Filed:

Dec 28, 2017

Appl. No.:

15/856424

Inventors:

- Menlo Park CA, US
Ashish Kumar Yadav - Mountain View CA, US
Daniel Dinu - Sunnyvale CA, US

International Classification:

H04L 29/08
G06F 15/18
G06F 17/30
G06Q 50/00

Abstract:

Systems, methods, and non-transitory computer readable media are configured to determine a likelihood of a rejection of a notification proposed for delivery to a recipient. A delivery determination for the notification can be performed. Subsequently, the notification can be delivered to the recipient based on the delivery determination.

US Patent:

20180107665, Apr 19, 2018

Filed:

Oct 17, 2016

Appl. No.:

15/295882

Inventors:

- Menlo Park CA, US
Daniel Dinu - Sunnyvale CA, US
Neal Suresh Vora - San Jose CA, US

International Classification:

G06F 17/30
G06Q 50/00
G06N 99/00

Abstract:

Systems, methods, and non-transitory computer-readable media according to certain aspects can obtain a goal associated with a page provided by a social networking system. Potential recommendations for the page can be determined based on a first machine learning model. The potential recommendations can be ranked based on a second machine learning model to identify a subset of recommendations relating to the goal.

US Patent:

20180103005, Apr 12, 2018

Filed:

Oct 10, 2016

Appl. No.:

15/289729

Inventors:

- Menlo Park CA, US
Komal Kapoor - Bellevue WA, US
Daniel Dinu - Sunnyvale CA, US
Bradley Ray Green - Snohomish WA, US
Naman Jain - Randolph NJ, US

International Classification:

H04L 12/58

Abstract:

Systems, methods, and non-transitory computer readable media are configured to receive values associated with features corresponding to an instance involving a page of a social networking system and an administrator of the page. The values associated with the features are applied to a machine learning model. A probability that the administrator of the page will take action on the page in response to receipt of an electronic notification provided to the administrator is determined based on the machine learning model.

US Patent:

20180089578, Mar 29, 2018

Filed:

Sep 26, 2016

Appl. No.:

15/276587

Inventors:

- Menlo Park CA, US
Daniel Dinu - Sunnyvale CA, US

International Classification:

G06N 99/00
H04L 29/08

Abstract:

Systems, methods, and non-transitory computer-readable media can train a machine learning model to classify at least one user account as either a first type of account or a second type of account based at least in part on one or more respective features corresponding to the user account and determine that a first user account that was created as the first type of account should be converted to the second type of account based at least in part on the machine learning model.

US Patent:

20170155907, Jun 1, 2017

Filed:

Nov 7, 2016

Appl. No.:

15/345315

Inventors:

- Redmond WA, US
Eric S. Christoffersen - Beaux Arts WA, US
Sean M. Callahan - St. Paul MN, US
Daniel Dinu - Redmond WA, US
Barry Friemel - Redmond WA, US
Weidong Zhao - Bellevue WA, US
Yongjun Wu - Bellevue WA, US
William Chen - Issaquah WA, US

Assignee:

Microsoft Technology Licensing, LLC - Redmond WA

International Classification:

H04N 19/13
H04N 19/105
H04N 19/43
H04N 19/15
H04N 19/593
H04N 19/82

Abstract:

Video decoding innovations for multithreading implementations and graphics processor unit (“GPU”) implementations are described. For example, for multithreaded decoding, a decoder uses innovations in the areas of layered data structures, picture extent discovery, a picture command queue, and/or task scheduling for multithreading. Or, for a GPU implementation, a decoder uses innovations in the areas of inverse transforms, inverse quantization, fractional interpolation, intra prediction using waves, loop filtering using waves, memory usage and/or performance-adaptive loop filtering. Innovations are also described in the areas of error handling and recovery, determination of neighbor availability for operations such as context modeling and intra prediction, CABAC decoding, computation of collocated information for direct mode macroblocks in B slices, reduction of memory consumption, implementation of trick play modes, and picture dropping for quality adjustment.