Abstract: Acoustic features are batched into two different batches. The second batch of the two batches is made in response to a detection of a word hypothesis output by a speech recognition network that received the first batch. The number of acoustic feature frames of the second batch is equal to a second batch size greater than the first batch size. The second batch is also to the speech recognition network for processing.

Abstract: Embodiments may include collection of a first batch of acoustic feature frames of an audio signal, the number of acoustic feature frames of the first batch equal to a first batch size, input of the first batch to a speech recognition network, collection, in response to detection of a word hypothesis output by the speech recognition network, of a second batch of acoustic feature frames of the audio signal, the number of acoustic feature frames of the second batch equal to a second batch size greater than the first batch size, and input of the second batch to the speech recognition network.

Abstract: Embodiments may include collection of a first batch of acoustic feature frames of an audio signal, the number of acoustic feature frames of the first batch equal to a first batch size, input of the first batch to a speech recognition network, collection, in response to detection of a word hypothesis output by the speech recognition network, of a second batch of acoustic feature frames of the audio signal, the number of acoustic feature frames of the second batch equal to a second batch size greater than the first batch size, and input of the second batch to the speech recognition network.

Abstract: Embodiments may include collection of a first batch of acoustic feature frames of an audio signal, the number of acoustic feature frames of the first batch equal to a first batch size, input of the first batch to a speech recognition network, collection, in response to detection of a word hypothesis output by the speech recognition network, of a second batch of acoustic feature frames of the audio signal, the number of acoustic feature frames of the second batch equal to a second batch size greater than the first batch size, and input of the second batch to the speech recognition network.

Abstract: Image registration is described. In an embodiment an image registration system executes automatic registration of images, for example medical images. In an example, semantic information is computed for each of the images to be registered comprising information about the types of objects in the images and the certainty of that information. In an example a mapping is found to register the images which takes into account the intensities of the image elements as well as the semantic information in a manner which is weighted by the certainty of that semantic information. For example, the semantic information is computed by estimating posterior distributions for the locations of anatomical structures by using a regression forest and transforming the posterior distributions into a probability map. In an example the mapping is found as a global point of inflection of an energy function, the energy function having a term related to the semantic information.

Abstract: Image-hosted data encryption implementations are presented that encrypt and decrypt data within a host image. A bit stream representing a data item and a host image are accessed. The host image has pixels which include one or more color channels that are represented by bits that are interpreted when the host image is rendered and other bits that are not interpreted when the host image is rendered. The bit stream is inserted into the host image by injecting bits into the bits of the one or more color channels not interpreted when the host image is rendered. In this way, the inserted data is not seen when the host image is rendered. Decryption involves extracting bits associated with a bit stream from the bits of the one or more color channels of the host image pixels not interpreted when the host image is rendered, and then reconstructing the bit stream.

Abstract: Image registration is described. In an embodiment an image registration system executes automatic registration of images, for example medical images. In an example, semantic information is computed for each of the images to be registered comprising information about the types of objects in the images and the certainty of that information. In an example a mapping is found to register the images which takes into account the intensities of the image elements as well as the semantic information in a manner which is weighted by the certainty of that semantic information. For example, the semantic information is computed by estimating posterior distributions for the locations of anatomical structures by using a regression forest and transforming the posterior distributions into a probability map. In an example the mapping is found as a global point of inflection of an energy function, the energy function having a term related to the semantic information.

Abstract: While performing prostate brachytherapy in which radioactive seeds are implanted into the prostate, both X-ray and transrectal ultrasound (TRUS) data are collected. The X-ray data indicate all of the implanted seeds, but not their relative position within the prostate. The TRUS data correspond to both axial and longitudinal ultrasound images. A subset of the implanted seeds can be automatically (or manually) detected in the TRUS data (using longitudinal ultrasound images), and the prostate is indicated in the TRUS data (using axial ultrasound images). Registration is performed between the X-ray and TRUS data, thereby identifying the disposition of all implanted seeds within the prostate, enabling dosimetry to be intraoperatively determined. A medical practitioner can thus determine whether and where to implant additional seeds to achieve a desired optimal dosage and do so before the prostate brachytherapy procedure is concluded.

Abstract: A system for image registration and quantitative feature extraction for multiple image sets. The system includes an imaging workstation having a data processor and memory in communication with a database server in communication. The a data processor capable of inputting and outputting data and instructions to peripheral devices and operating pursuant to a software product and accepts instructions from a graphical user interface capable of interfacing with and navigating the imaging software product. The imaging software product is capable of instructing the data processor, to register images, segment images and to extract features from images and provide instructions to store and retrieve one or more registered images, segmented images, quantitative image features and quantitative image data and from the database server.

Abstract: A system for image registration and quantitative feature extraction for multiple image sets. The system includes an imaging workstation having a data processor and memory in communication with a database server in communication. The a data processor capable of inputting and outputting data and instructions to peripheral devices and operating pursuant to a software product and accepts instructions from a graphical user interface capable of interfacing with and navigating the imaging software product. The imaging software product is capable of instructing the data processor, to register images, segment images and to extract features from images and provide instructions to store and retrieve one or more registered images, segmented images, quantitative image features and quantitative image data and from the database server.

Abstract: An ultrasound system automatically measures fetal head size from ultrasound images. An ultrasound image of the fetal head is detected. A radial maxima point is identified on each of a plurality of radii extending from a substantially common vertex point within the fetal head image. Each radial maxima point corresponds to an ultrasound sample along its corresponding radius, and has a maximum ultrasound echo strength. Outlier points are removed and the curve filtered to derive an initial fetal head boundary. An inner fetal head boundary and outer fetal head boundary are derived from the initial fetal head boundary and a predetermined fetal skull thickness, and fetal head size is computed from the inner fetal head boundary and the outer fetal head boundary. Processing is allocated among multiprocessors and performed in pipeline fashion to enable real-time interactive imaging and measuring.