Facebook, Inc.

Chien-Nan Chen

Evgeny V Kuzyakov

David Young Joon Pio

Renbin Peng

Systems, methods, and non-transitory computer-readable media can provide at least one frame of a content item to a saliency prediction model, the saliency prediction model being trained to identify salient points of interest that appear in content items. Information describing at least a first salient point of interest that appears in the at least one frame can be obtained from the saliency prediction model. The first salient point of interest can be predicted to be of interest to users accessing the content item. A view-based projection can be applied to a region corresponding to the first salient point of interest, wherein the view-based projection enhances a quality in which the region is presented.

Systems and methods for provisioning content

Systems, methods, and non-transitory computer-readable media can obtain a spherical media content item that captures at least one scene from a plurality of different positions. A three-dimensional object with a plurality of faces that converge at a point is determined. At least a first portion of the spherical media content item is mapped to at least a first face in the plurality of faces.

Systems and methods for content streaming

Systems, methods, and non-transitory computer-readable media can obtain a spherical media content item that captures at least one scene from a plurality of different positions. A three-dimensional shape having a plurality of faces is determined, each face being associated with a respective set of original boundaries. A respective set of expanded boundaries for one or more of the plurality of faces is determined. Respective portions of the spherical media content item are mapped to each of the one or more faces, wherein a first portion of the spherical media content item is mapped to a first face having a set of expanded boundaries.

Systems, methods, and non-transitory computer-readable media can determine a first position corresponding to a user focal point during presentation of a given scene of a content item. A celestial representation of the given scene is determined. The celestial representation identifies one or more audio-based points of interest in the scene. A second position corresponding to the user focal point is determined based at least in part on the celestial representation and the first position. The second position is predicted to be the position of the user focal point during presentation of the scene.

Systems and methods for presenting content

Systems, methods, and non-transitory computer-readable media can determine at least one salient point of interest in a frame of a content item based at least in part on a saliency prediction model, the saliency prediction model being trained to identify salient points of interest that appear in content items; determine a barrel projection representation for the frame; and apply a view-based projection to the barrel projection representation for the frame, wherein the view-based projection enhances a quality in which a region corresponding to the at least one salient point of interest is presented.

Systems and methods for provisioning content using barrel projection representation

Systems, methods, and non-transitory computer-readable media can determine a first position corresponding to a user focal point prior to presenting a given frame of a content item. A landscape representation of the given frame is determined. The landscape representation describes the respective popularity of one or more regions in the frame as a topology. A second position corresponding to the user focal point is determined based at least in part on the landscape representation and the first position. The second position is predicted to be the position of the user focal point when presenting the given frame.

Growing community of inventors

Champaign, IL, United States of America

Chien-Nan Chen