Jovan O Popovic, 51Seattle, WA

2013012, May 16, 2013

Feb 26, 2010

12/714028

Jovan Popovic - Seattle WA, US
Jen-Chan Chien - Saratoga CA, US
Chintan Intwala - Santa Clara CA, US
Sarah A. Kong - Cupertino CA, US

G09G 5/00

345647, 345650, 345661, 345672

Methods and apparatus for manipulating digital images. A warping module is described that enables the manipulation of a surface by selectively deforming portions of the surface while maintaining local rigidity. The user may position multiple control points on a surface to constrain deformation. The user may specify multiple properties (e.g., translation, rotation, depth, and scale) at each control point. A mesh may be overlaid on the surface. The warping module may perform an initialization in which the properties are propagated other vertices in the mesh to generate an initial deformed mesh. The warping module may then perform an iterative optimization operation on the deformed mesh to improve the deformation while retaining local rigidity. Thus, instead of moving every pixel in the surface, the warping module moves or adjusts coordinates of the vertices of the mesh. The surface is then deformed according to the deformed mesh.

2013012, May 23, 2013

Nov 30, 2010

12/956971

Jovan Popovic - Seattle WA, US
Sergey V. Levine - Redmond WA, US

G06T 13/80

345473

An interactive application may include a quasi-physical simulator configured to determine the configuration of animated characters as they move within the application and are acted on by external forces. The simulator may work together with a parameterized animation module that synthesizes and provides reference poses for the animation from example motion clips that it has segmented and parameterized. The simulator may receive input defining a trajectory for an animated character and input representing one or more external forces acting on the character, and may perform a quasi-physical simulation to determine a pose for the character in the current animation frame in reaction to the external forces. The simulator may enforce a goal constraint that the animated character follows the trajectory, e.g., by adding a non-physical force to the simulation, the magnitude of which may be dependent on a torque objective that attempts to minimize the use of such non-physical forces.

5966140, Oct 12, 1999

Jun 20, 1997

8/880090

Jovan Popovic - Pittsburgh PA
Hugues H. Hoppe - Seattle WA

Microsoft Corporation - Redmond WA

G06F 1500

345441

A method for creating progressive simplicial complexes (PSC), including a new format for storing and transmitting arbitrary geometric models for computer graphics is presented. A PSC captures a graphical model as a coarse base model together with a sequence of refinement transformations that progressively recover detail. The PSC method uses general refinement transformations, allowing the given model to be an arbitrary triangulation of a complex shape, and the base model to be a single vertex. The PSC model defines a continuous sequence of approximating models for run-time level-of-detail control, allows smooth transitions between any pair of models in the sequence, supports compression, progressive transmission on computer network like the Internet or an intranet, and offers a space-efficient representation of an arbitrary geometric model.

2022030, Sep 22, 2022

Jun 8, 2022

17/805907

- SAN JOSE CA, US
Cristin Ailidh Fraser - San Diego CA, US
Aseem Agarwala - Seattle WA, US
Lubomira Dontcheva - Seattle WA, US
Joel Richard Brandt - Venice CA, US
Jovan Popovic - Seattle WA, US

G06T 7/12
G06T 7/162
G11B 27/10
G06V 20/40
G06V 40/16

Embodiments are directed to video segmentation based on detected video features. More specifically, a segmentation of a video is computed by determining candidate boundaries from detected feature boundaries from one or more feature tracks; modeling different segmentation options by constructing a graph with nodes that represent candidate boundaries, edges that represent candidate segments, and edge weights that represent cut costs; and computing the video segmentation by solving a shortest path problem to find the path through the edges (segmentation) that minimizes the sum of edge weights along the path (cut costs). A representation of the video segmentation is presented, for example, using interactive tiles or a video timeline that represent(s) the video segments in the segmentation.

2022030, Sep 22, 2022

Jun 2, 2022

17/805076

- SAN JOSE CA, US
Xue Bai - Bellevue WA, US
Aseem Agarwala - Seattle WA, US
Joel R. Brandt - Venice CA, US
Jovan Popovic - Seattle WA, US
Lubomira Dontcheva - Seattle WA, US
Joy Oakyung Kim - Menlo Park CA, US
Seth Walker - Oakland CA, US

G06V 20/40
G06K 9/62
G11B 27/19
G11B 27/00
G10L 25/78

Embodiments are directed to segmentation and hierarchical clustering of video. In an example implementation, a video is ingested to generate a multi-level hierarchical segmentation of the video. In some embodiments, the finest level identifies a smallest interaction unit of the video—semantically defined video segments of unequal duration called clip atoms. Clip atom boundaries are detected in various ways. For example, speech boundaries are detected from audio of the video, and scene boundaries are detected from video frames of the video. The detected boundaries are used to define the clip atoms, which are hierarchically clustered to form a multi-level hierarchical representation of the video. In some cases, the hierarchical segmentation identifies a static, pre-computed, hierarchical set of video segments, where each level of the hierarchical segmentation identifies a complete set (i.e., covering the entire range of the video) of disjoint (i.e., non-overlapping) video segments with a corresponding level of granularity.

2022029, Sep 15, 2022

Jun 2, 2022

17/805075

- SAN JOSE CA, US
Xue Bai - Bellevue WA, US
Aseem Agarwala - Seattle WA, US
Joel R. Brandt - Venice CA, US
Jovan Popovic - Seattle WA, US
Lubomira Dontcheva - Seattle WA, US
Joy Oakyung Kim - Menlo Park CA, US
Seth Walker - Oakland CA, US

G06V 20/40
G06K 9/62
G11B 27/19
G11B 27/00
G10L 25/78

Embodiments are directed to segmentation and hierarchical clustering of video. In an example implementation, a video is ingested to generate a multi-level hierarchical segmentation of the video. In some embodiments, the finest level identifies a smallest interaction unit of the video—semantically defined video segments of unequal duration called clip atoms. Clip atom boundaries are detected in various ways. For example, speech boundaries are detected from audio of the video, and scene boundaries are detected from video frames of the video. The detected boundaries are used to define the clip atoms, which are hierarchically clustered to form a multi-level hierarchical representation of the video. In some cases, the hierarchical segmentation identifies a static, pre-computed, hierarchical set of video segments, where each level of the hierarchical segmentation identifies a complete set (i.e., covering the entire range of the video) of disjoint (i.e., non-overlapping) video segments with a corresponding level of granularity.

2022029, Sep 15, 2022

Jun 2, 2022

17/805080

- SAN JOSE CA, US
Xue Bai - Bellevue WA, US
Aseem Agarwala - Seattle WA, US
Joel R. Brandt - Venice CA, US
Jovan Popovic - Seattle WA, US
Lubomira Dontcheva - Seattle WA, US
Joy Oakyung Kim - Menlo Park CA, US
Seth Walker - Oakland CA, US

G06V 20/40
G06K 9/62
G11B 27/19
G11B 27/00
G10L 25/78

Embodiments are directed to segmentation and hierarchical clustering of video. In an example implementation, a video is ingested to generate a multi-level hierarchical segmentation of the video. In some embodiments, the finest level identifies a smallest interaction unit of the video—semantically defined video segments of unequal duration called clip atoms. Clip atom boundaries are detected in various ways. For example, speech boundaries are detected from audio of the video, and scene boundaries are detected from video frames of the video. The detected boundaries are used to define the clip atoms, which are hierarchically clustered to form a multi-level hierarchical representation of the video. In some cases, the hierarchical segmentation identifies a static, pre-computed, hierarchical set of video segments, where each level of the hierarchical segmentation identifies a complete set (i.e., covering the entire range of the video) of disjoint (i.e., non-overlapping) video segments with a corresponding level of granularity.

2020029, Sep 17, 2020

May 29, 2020

16/887418

- San Jose CA, US
Jovan Popovic - Seattle WA, US
Deepali Aneja - Seattle WA, US
David Simons - Seattle WA, US

G10L 15/197
G06N 3/04
G06N 3/08
G10L 15/02
G10L 15/06
G10L 21/0316
G10L 25/21
G10L 25/24

Disclosed systems and methods predict visemes from an audio sequence. In an example, a viseme-generation application accesses a first audio sequence that is mapped to a sequence of visemes. The first audio sequence has a first length and represents phonemes. The application adjusts a second length of a second audio sequence such that the second length equals the first length and represents the phonemes. The application adjusts the sequence of visemes to the second audio sequence such that phonemes in the second audio sequence correspond to the phonemes in the first audio sequence. The application trains a machine-learning model with the second audio sequence and the sequence of visemes. The machine-learning model predicts an additional sequence of visemes based on an additional sequence of audio.