Abstract: Apparatuses, systems, and techniques for texture synthesis from small input textures in images using convolutional neural networks. In at least one embodiment, one or more convolutional layers are used in conjunction with one or more transposed convolution operations to generate a large textured output image from a small input textured image while preserving global features and texture, according to various novel techniques described herein.

Abstract: Apparatuses, systems, and techniques to generate interpolated video frames. In at least one embodiment, an interpolated video frame is generated based, at least in part, on a first set of pixel data sampled from a first video frame, and a second set of pixel data sampled from a second video frame based, at least in part, on a set of forward pointing motion vectors from the first video frame to the second video frame.

Abstract: Apparatuses, systems, and techniques to generate interpolated video frames. In at least one embodiment, an interpolated video frame is generated based, at least in part, on one of a plurality of possible motions of one or more objects from a first video frame to a second video frame.

Abstract: Apparatuses, systems, and techniques for texture synthesis from small input textures in images using convolutional neural networks. In at least one embodiment, one or more convolutional layers are used in conjunction with one or more transposed convolution operations to generate a large textured output image from a small input textured image while preserving global features and texture, according to various novel techniques described herein.

Abstract: Apparatuses, systems, and techniques to enhance video. In at least one embodiment, one or more neural networks are used to create, from a first video, a second video having a higher frame rate, higher resolution, or reduced number of missing or corrupt video frames.

Abstract: A neural network architecture is disclosed for performing video frame prediction using a sequence of video frames and corresponding pairwise optical flows. The neural network processes the sequence of video frames and optical flows utilizing three-dimensional convolution operations, where time (or multiple video frames in the sequence of video frames) provides the third dimension in addition to the two-dimensional pixel space of the video frames. The neural network generates a set of parameters used to predict a next video frame in the sequence of video frames by sampling a previous video frame utilizing spatially-displaced convolution operations. In one embodiment, the set of parameters includes a displacement vector and at least one convolution kernel per pixel. Generating a pixel value in the next video frame includes applying the convolution kernel to a corresponding patch of pixels in the previous video frame based on the displacement vector.

Abstract: A neural network architecture is disclosed for performing image in-painting using partial convolution operations. The neural network processes an image and a corresponding mask that identifies holes in the image utilizing partial convolution operations, where the mask is used by the partial convolution operation to zero out coefficients of the convolution kernel corresponding to invalid pixel data for the holes. The mask is updated after each partial convolution operation is performed in an encoder section of the neural network. In one embodiment, the neural network is implemented using an encoder-decoder framework with skip links to forward representations of the features at different sections of the encoder to corresponding sections of the decoder.

Abstract: A method for automatic segmentation of intra-cochlear anatomy in post-implantation CT image of bilateral cochlear implant recipients includes coarsely segmenting a labyrinth with a labyrinth surface chosen from a library of inner ear anatomy shapes; creating a target specific ASM for each of the labyrinth and the SOIs using a set of inner ear anatomy surfaces selected from the library of inner ear anatomy shapes such that the set of inner ear anatomy surfaces has the smallest dissimilarity quantity with the coarsely localized labyrinth surface in the post-implantation CT image; refining the coarsely segmented labyrinth surface by performing an ASM-based segmentation of the labyrinth using the target-specific ASM of the labyrinth to obtain a segmented labyrinth; and fitting the points of the target-specific ASM of the SOIs to their corresponding points on the segmented labyrinth to segment the SOIs in the post-implantation CT image.

Abstract: A method for automatic segmentation of intra-cochlear anatomy in post-implantation CT image of bilateral cochlear implant recipients includes coarsely segmenting a labyrinth with a labyrinth surface chosen from a library of inner ear anatomy shapes; creating a target specific ASM for each of the labyrinth and the SOIs using a set of inner ear anatomy surfaces selected from the library of inner ear anatomy shapes such that the set of inner ear anatomy surfaces has the smallest dissimilarity quantity with the coarsely localized labyrinth surface in the post-implantation CT image; refining the coarsely segmented labyrinth surface by performing an ASM-based segmentation of the labyrinth using the target-specific ASM of the labyrinth to obtain a segmented labyrinth; and fitting the points of the target-specific ASM of the SOIs to their corresponding points on the segmented labyrinth to segment the SOIs in the post-implantation CT image.

Abstract: A method for automatic segmentation of intra-cochlear anatomy of a patient. The patient has an implanted ear and a normal contralateral ear. At least one computed tomography (CT) image is obtained to generate a first image corresponding to the normal contralateral ear and a second image corresponding to the implanted ear. Intra-cochlear surfaces of at least one first structure of interest (SOI) of the normal contralateral ear in the first image are segmented using at least one active shape model (ASM). Next, the segmented intra-cochlear surfaces in the first image is projected to the second image using a transformation function, thereby obtaining projected segmented intra-cochlear surfaces for the implanted ear in the second image.

Abstract: A method for automatic segmentation of intra-cochlear anatomy of a patient. The patient has an implanted ear and a normal contralateral ear. At least one computed tomography (CT) image is obtained to generate a first image corresponding to the normal contralateral ear and a second image corresponding to the implanted ear. Intra-cochlear surfaces of at least one first structure of interest (SOI) of the normal contralateral ear in the first image are segmented using at least one active shape model (ASM). Next, the segmented intra-cochlear surfaces in the first image is projected to the second image using a transformation function, thereby obtaining projected segmented intra-cochlear surfaces for the implanted ear in the second image.