Abstract: Apparatuses, systems, and techniques are presented to generate images representing realistic motion or activity. In at least one embodiment, one or more neural networks are used to select a first neural network to perform a first task based, at least in part, upon a performance estimated by a second neural network.

Abstract: Described herein are means for generating source models for transfer learning to application specific models used in the processing of medical imaging.

Abstract: Methods, systems, and media for segmenting images are provided. In some embodiments, the method comprises: generating an aggregate U-Net comprised of a plurality of U-Nets, wherein each U-Net in the plurality of U-Nets has a different depth, wherein each U-Net is comprised of a plurality of nodes Xi,j, wherein i indicates a down-sampling layer the U-Net, and wherein j indicates a convolution layer of the U-Net; training the aggregate U-Net by: for each training sample in a group of training samples, calculating, for each node in the plurality of nodes Xi,j, a feature map xi,j, wherein xi,j is based on a convolution operation performed on a down-sampling of an output from Xi?1,j when j=0, and wherein xi,j is based on a convolution operation performed on an up-sampling operation of an output from Xi+1,j?1 when j>0; and predicting a segmentation of a test image using the trained aggregate U-Net.

Abstract: Described herein are means for implementing fixed-point image-to-image translation using improved Generative Adversarial Networks (GANs). For instance, an exemplary system is specially configured for implementing a new framework, called a Fixed-Point GAN, which improves upon prior known methodologies by enhancing the quality of the images generated through global, local, and identity transformation. The Fixed-Point GAN as introduced and described herein, improves many applications dependant on image-to-image translation, including those in the field of medical image processing for the purposes of desease detection and localization. Other related embodiments are disclosed.

Abstract: Methods, systems, and media for discriminating and generating translated images are provided. In some embodiments, the method comprises: identifying a set of training images, wherein each image is associated with at least one domain from a plurality of domains; training a generator network to generate: i) a first fake image that is associated with a first domain; and ii) a second fake image that is associated with a second domain; training a discriminator network, using as inputs to the discriminator network: i) an image from the set of training images; ii) the first fake image; and iii) the second fake image; and using the generator network to generate, for an image not included in the set of training images at least one of: i) a third fake image that is associated with the first domain; and ii) a fourth fake image that is associated with the second domain.

Abstract: Disclosed are provided systems, methods, and apparatuses for implementing a multi-resolution neural network for use with imaging intensive applications including medical imaging. For example, a system having means to execute a neural network model formed from a plurality of layer blocks including an encoder layer block which precedes a plurality of decoder layer blocks includes: means for associating a resolution value with each of the plurality of layer blocks; means for processing via the encoder layer block a respective layer block input including a down-sampled layer block output processing, via decoder layer blocks, a respective layer block input including an up-sampled layer block output and a layer block output of a previous layer block associated with a prior resolution value of a layer block which precedes the respective decoder layer block; and generating the respective layer block output by summing or concatenating the processed layer block inputs.

Abstract: Methods, systems, and media for segmenting images are provided. In some embodiments, the method comprises: generating an aggregate U-Net comprised of a plurality of U-Nets, wherein each U-Net in the plurality of U-Nets has a different depth, wherein each U-Net is comprised of a plurality of nodes Xi,j, wherein i indicates a down-sampling layer the U-Net, and wherein j indicates a convolution layer of the U-Net; training the aggregate U-Net by: for each training sample in a group of training samples, calculating, for each node in the plurality of nodes Xi,j, a feature map xi,j, wherein xi,j is based on a convolution operation performed on a down-sampling of an output from Xi?1,j when j=0, and wherein xi,j is based on a convolution operation performed on an up-sampling operation of an output from Xi+1,j?1 when j>0; and predicting a segmentation of a test image using the trained aggregate U-Net.

Abstract: Methods, systems, and media for discriminating and generating translated images are provided. In some embodiments, the method comprises: identifying a set of training images, wherein each image is associated with at least one domain from a plurality of domains; training a generator network to generate: i) a first fake image that is associated with a first domain; and ii) a second fake image that is associated with a second domain; training a discriminator network, using as inputs to the discriminator network: i) an image from the set of training images; ii) the first fake image; and iii) the second fake image; and using the generator network to generate, for an image not included in the set of training images at least one of: i) a third fake image that is associated with the first domain; and ii) a fourth fake image that is associated with the second domain.

Abstract: Disclosed are provided systems, methods, and apparatuses for implementing a multi-resolution neural network for use with imaging intensive applications including medical imaging. For example, a system having means to execute a neural network model formed from a plurality of layer blocks including an encoder layer block which precedes a plurality of decoder layer blocks includes: means for associating a resolution value with each of the plurality of layer blocks; means for processing via the encoder layer block a respective layer block input including a down-sampled layer block output processing, via decoder layer blocks, a respective layer block input including an up-sampled layer block output and a layer block output of a previous layer block associated with a prior resolution value of a layer block which precedes the respective decoder layer block; and generating the respective layer block output by summing or concatenating the processed layer block inputs.

Abstract: A system and method of conference messaging between Universal Plug and Play (UPnP) telephony devices and Wide Area Network (WAN) devices is where a conference messaging session with WAN devices is initiated by at least one Telephony Control Point (TCP) via a session request which includes session information, media capabilities of at least one TCP, a list of WAN devices, and a subject of the conference messaging session. The conference messaging session is then established between the at least one TCP and the WAN devices by a Telephony Server (TS) in an UPnP telephony based home network based on the session request.

Abstract: A method and apparatus for restricting disclosure of information about a network device during a Remote Access (RA) service are provided, in which a virtual device is activated by an RA service device, a device or service for which information is to be disclosed to an external network is determined from among devices discovered by the RA service device and registered as an embedded device or service of the virtual device, a device description of the virtual device including the embedded device or service is dynamically generated, a Uniform Resource Locator (URL) of the RA service device, from which the device description of the virtual device is acquired, is generated and opened to a device of the other party which has remotely accessed the RA service device, and the device description of the virtual device is provided, upon receipt of a request for the device description of the virtual device through the URL.

Abstract: Various embodiments relate to a cloud-based temporary app management system. In one aspect of the invention, a cloud service provides a temporary app from a vendor. The temporary app is a mobile application that is temporarily accessible on a mobile device when a designated event (e.g. a change in the location or surroundings of the mobile device, a particular type of data stored in the mobile device, etc.) is detected by a mobile device. The cloud service receives event information from the mobile device. The event information indicates an event detected by the mobile device. The cloud service then determines whether to make the temporary app accessible on the mobile device based on an analysis of the designated event and the event information received from the mobile device.

Abstract: A method for power management comprises registering and subscribing one or more endpoint devices to a remote power management service. The one or more endpoint devices are connected with a local network. Power state information of the one or more endpoint devices is obtained. Power state of one or more endpoint devices is changed using the remote power management service.

Abstract: In a first embodiment of the present invention, a method for operating a presence server in a home network is provided, the method comprising: receiving a request for presence information; sending an event notification to all subscribed control points informing them of the request for presence information; receiving an action from one of the subscribed control points accepting or rejecting the request for presence information; and if the action received from the one of the subscribed control points accepts the request for presence information, causing presence information regarding the one of the subscribed control points to be sent to the entity that sent the request for presence information.

Abstract: In one embodiment, a method for providing voice communications in a packet switched network protocol through a home network is provided, the method comprising: receiving, at a first home network device, an incoming call in the packet switched network protocol; notifying a second home network device of the incoming call; receiving an indication from the second home network device that the second home network device accepts the call; and forwarding the incoming call to the second home network device.

Abstract: A method for operating a telephony server in a home network is provided, comprising: receiving a first registration action from a first telephony control point in the home network, the first registration action including an identification of a first user within a first telephony control point; supplying a first authentication identifier to the first telephony control point, wherein the first authentication identifier is bound to both the first telephony control point and the first user; receiving an invoked action granting a telephony action from the first telephony control point, wherein the invoked action includes an authentication identifier; determining if the authentication identifier received in the invoked action matches the authentication identifier bound to the user which the telephony action involves; and executing the telephony action if the authentication information received in the invoked action matches the authentication identifier bound to the user which the telephony action involves.

Abstract: Various embodiments relate to a cloud-based temporary app management system. In one aspect of the invention, a cloud service provides a temporary app from a vendor. The temporary app is a mobile application that is temporarily accessible on a mobile device when a designated event (e.g. a change in the location or surroundings of the mobile device, a particular type of data stored in the mobile device, etc.) is detected by a mobile device. The cloud service receives event information from the mobile device. The event information indicates an event detected by the mobile device. The cloud service then determines whether to make the temporary app accessible on the mobile device based on an analysis of the designated event and the event information received from the mobile device.

Abstract: A method for power management comprises registering and subscribing one or more endpoint devices to a remote power management service. The one or more endpoint devices are connected with a local network. Power state information of the one or more endpoint devices is obtained. Power state of one or more endpoint devices is changed using the remote power management service.

Abstract: In a first embodiment of the present invention, a method for remoting user/callee presence information in a home network is provided, the method comprising: subscribing, from a home network presence client, to one or more home network presence servers, for presence information relating to one or more users; receiving, from one of the home network presence servers, a change in presence information relating to one of said one or more users; and notifying a user of a device associated with the home network presence client of the change in presence information.

Abstract: In one embodiment, a method for providing voice communications in a packet switched network protocol through a home network is provided, the method comprising: receiving, at a first home network device, an incoming call in the packet switched network protocol; notifying a second home network device of the incoming call; receiving an indication from the second home network device that the second home network device accepts the call; and forwarding the incoming call to the second home network device.