Abstract: A method of rendering a stereoscopic 3D image from a single image, including receiving a collection of pairs of 3D images including an input image and an output image that is a 3D pair of the input image, training a neural network composed of convolutional layers without any fully connected layers, with pairs of 3D images from the collection, to receive an input image and generate an output image that is a 3D pair of the input image, wherein the neural network is provided as an application on a computing device, receiving an input image and generating an output image that is a 3D pair of the input image by the neural network.

Abstract: A method of rendering a stereoscopic 3D image from a single image, including receiving a collection of pairs of 3D images including an input image and an output image that is a 3D pair of the input image, training a neural network composed of convolutional layers without any fully connected layers, with pairs of 3D images from the collection, to receive an input image and generate an output image that is a 3D pair of the input image, wherein the neural network is provided as an application on a computing device, receiving an input image and generating an output image that is a 3D pair of the input image by the neural network.

Abstract: A method and system for creating a depth signature from plural images for providing watermark information related to the images. The method comprises analyzing a pair of images, each image containing a plurality of elements, identifying a first element in one of the pair of images, identifying plural elements in the other of the pair of images. The method further comprises measuring a disparity parameter between the first element and a set of the plural elements, matching the first element from the set of plural elements, the matched second element having the smallest measured disparity parameter, and computing a signature based at least in part on the measured disparity between the first and second elements.

Abstract: A system and method for multimedia content protection on elastic cloud infrastructures is presented. The system can be used to protect various multi-media contents, including regular 2D videos, new 3D videos, animated graphics, images, audios clips, songs, and music clips. The system can run on private clouds, public clouds, or any combination of public-private clouds. The system is scalable and cost effective.

Abstract: A method to design, implement and create distributed indexes for storing and comparing fingerprints of videos is presented. The method effectively utilizes cloud computing platforms that offer varying amounts of computing resources. The method enables the distributed index to scale to large numbers of data points and the distributed index is robust to failures within the computing resources maintaining the index. The method minimizes the memory required to maintain the distributed index and reduces the I/O operations needed to process operations performed on the index. The method improves the efficiency of the index to process queries.

Abstract: A graphics processing unit controller configured to be communicatively coupled to one or more graphics processing units and one or more virtual machines, comprising: a unit collection module which is configured to store information regarding the available resources of one or more graphics processing units; and a scheduler module configured to allocate at least part of the one or more graphics processing units to the execution of a compute kernel in response to receipt of a request for the execution of the compute kernel, wherein the scheduler is configured to allocate at least part of the one or more graphics processing units based on a comparison between the available resources of the one or more graphics processing units and required resources for the execution of the compute kernel.

Abstract: A computer-implemented method for detecting a copy of a reference video, comprises segmenting respective ones of multiple frames of the reference video into multiple regions, determining sets of image features appearing in respective ones of the multiple frames, determining a measure for the relative number of image features for a given region across the multiple frames, generating a spatio-temporal signature for the reference video using the determined measures, and comparing the signature for the reference video against a spatio-temporal signature of a query video to determine a likelihood of a match.

Abstract: Apparatus, comprising a wireless transceiver to wirelessly communicate with multiple recipients, control logic coupled to the wireless transceiver to determine an amount of available bandwidth for multicasting multiple data streams for the recipients, the control logic to select an encoded data stream including data substreams relating to at least first and second video reference views and corresponding depth data for respective ones of the video reference views to transmit to a recipient via the wireless transceiver on the basis of the determined bandwidth.

Abstract: A method of transmitting a live video stream and a pre-recorded video stream over a wireless network to a mobile device, the method including: receiving the live video stream and the pre-recorded video stream for a current scheduling window; dividing the live video stream and the pre-recorded video stream into segments in the current scheduling window, each segment including an aggregate data amount, a transmission start time and a transmission end time; determining a transmission schedule for the current scheduling window, said transmission schedule including the segments; locating slack time slots, the slack time slots being time slots for which no segment is allocated; and scheduling bursts associated with the pre-recorded video stream in the slack time slots.

Abstract: A system and method for multimedia content protection on elastic cloud infrastructures is presented. The system can be used to protect various multi-media contents, including regular 2D videos, new 3D videos, animated graphics, images, audios clips, songs, and music clips. The system can run on private clouds, public clouds, or any combination of public-private clouds. The system is scalable and cost effective.

Abstract: A graphics processing unit controller, host system, and methods A graphics processing unit controller configured to be communicatively coupled to one or more graphics processing units and one or more virtual machines, the controller comprising: a scheduler module configured to allocate at least part of one or more graphics processing units to the execution of a compute kernel in response to receipt of a request for the execution of the compute kernel during the running of an application by a virtual machine.

Abstract: A graphics processing unit controller configured to be communicatively coupled to one or more graphics processing units and one or more virtual machines, comprising: a unit collection module which is configured to store information regarding the available resources of one or more graphics processing units; and a scheduler module configured to allocate at least part of the one or more graphics processing units to the execution of a compute kernel in response to receipt of a request for the execution of the compute kernel, wherein the scheduler is configured to allocate at least part of the one or more graphics processing units based on a comparison between the available resources of the one or more graphics processing units and required resources for the execution of the compute kernel.

Abstract: A method to design, implement and create distributed indexes for storing and comparing fingerprints of videos is presented. The method effectively utilises cloud computing platforms that offer varying amounts of computing resources. The method enables the distributed index to scale to large numbers of data points and the distributed index is robust to failures within the computing resources maintaining the index. The method minimizes the memory required to maintain the distributed index and reduces the I/O operations needed to process operations performed on the index. The method improves the efficiency of the index to process queries.

Abstract: A method and system for creating a depth signature from plural images for providing watermark information related to the images. The method comprises analysing a pair of images, each image containing a plurality of elements, identifying a first element in one of the pair of images, identifying plural elements in the other of the pair of images. The method further comprises measuring a disparity parameter between the first element and a set of the plural elements, matching the first element from the set of plural elements, the matched second element having the smallest measured disparity parameter, and computing a signature based at least in part on the measured disparity between the first and second elements.

Abstract: A method for processing at least part of a large-scale dataset, the method comprising: receiving a dataset including a plurality of data points; generating a hash value for at least some of the data points; sorting the generated hash values into a plurality of buckets of identical or substantially identical hash values; generating a similarity matrix for each of the buckets; and applying a machine learning algorithm to the similarity matrices.

Abstract: In accordance with the embodiments of the present invention, a method and engine for assigning semantic tags to segments within media. The invention receives media and extracts textual information related to the media's content. It processes the textual information and creates a list of topics related to the content. The invention segments the media and intelligently assigns topical tags to the segments. The semantically segmented media data is outputted for storage or analysis.

Abstract: A method of transmitting a live video stream and a pre-recorded video stream over a wireless network to a mobile device, the method including: receiving the live video stream and the pre-recorded video stream for a current scheduling window; dividing the live video stream and the pre-recorded video stream into segments in the current scheduling window, each segment including an aggregate data amount, a transmission start time and a transmission end time; determining a transmission schedule for the current scheduling window, said transmission schedule including the segments; locating slack time slots, the slack time slots being time slots for which no segment is allocated; and scheduling bursts associated with the pre-recorded video stream in the slack time slots.

Abstract: Apparatus, comprising a wireless transceiver to wirelessly communicate with multiple recipients, control logic coupled to the wireless transceiver to determine an amount of available bandwidth for multicasting multiple data streams for the recipients, the control logic to select an encoded data stream including data substreams relating to at least first and second video reference views and corresponding depth data for respective ones of the video reference views to transmit to a recipient via the wireless transceiver on the basis of the determined bandwidth.

Abstract: A computer-implemented method for detecting a copy of a reference video, comprises segmenting respective ones of multiple frames of the reference video into multiple regions, determining sets of image features appearing in respective ones of the multiple frames, determining a measure for the relative number of image features for a given region across the multiple frames, generating a spatio-temporal signature for the reference video using the determined measures, and comparing the signature for the reference video against a spatio-temporal signature of a query video to determine a likelihood of a match.

Abstract: In accordance with the embodiments of the present invention, a method and engine for assigning semantic tags to segments within media. The invention receives media and extracts textual information related to the media's content. It processes the textual information and creates a list of topics related to the content. The invention segments the media and intelligently assigns topical tags to the segments. The semantically segmented media data is outputted for storage or analysis.