Abstract: A system includes a computing platform having a hardware processor and a memory storing a software code and a neural network (NN) having multiple layers including a last activation layer and a loss layer. The hardware processor executes the software code to identify different combinations of layers for testing the NN, each combination including candidate function(s) for the last activation layer and candidate function(s) for the loss layer. For each different combination, the software code configures the NN based on the combination, inputs, into the configured NN, a training dataset including multiple data objects, receives, from the configured NN, a classification of the data objects, and generates a performance assessment for the combination based on the classification. The software code determines a preferred combination of layers for the NN including selected candidate functions for the last activation layer and the loss layer, based on a comparison of the performance assessments.

Abstract: The present disclosure relates to a test device to be inserted into the syringe receptacle of a pump, for example for a blood treatment apparatus, for testing at least one function of the pump and/or of the blood treatment apparatus. The test device includes a holding device for releasably holding the test device on the housing of the blood treatment apparatus. In addition, the test device includes at least one movable actuator and at least one electric motor arranged to directly or indirectly move said movable actuator or a section thereof.

Abstract: The present disclosure relates to a bag holder for receiving a first effluent bag, which serves for receiving effluent resulting from a blood treatment being performed by a blood treatment apparatus and includes at least one closable effluent outlet opening and at least one effluent inlet opening. The bag holder includes one or several rollers for rolling the bag holder and at least one of the following features: a holding device for holding the first effluent bag in the bag holder, a receptacle for receiving the first effluent bag or a section thereof, a pump section for conveying effluent out of first effluent bag being received in the bag holder, at least one effluent outlet line for guiding effluent towards the pump section and/or out of the first effluent bag and/or at least one basin line for guiding effluent into the first effluent bag and/or away from the bag holder.

Abstract: One embodiment of the present invention sets forth a technique for performing frame interpolation. The technique includes generating (i) a first set of feature maps based on a first set of rendering features associated with a first key frame, (ii) a second set of feature maps based on a second set of rendering features associated with a second key frame, and (iii) a third set of feature maps based on a third set of rendering features associated with a target frame. The technique also includes applying one or more neural networks to the first, second, and third set of feature maps to generate a set of mappings from a first set of pixels in the first key frame to a second set of pixels in the target frame. The technique further includes generating the target frame based on the set of mappings.

Abstract: Various embodiments set forth systems and techniques for changing a face within an image. The techniques include receiving a first image including a face associated with a first facial identity; generating, via a machine learning model, at least a first texture map and a first position map based on the first image; rendering a second image including a face associated with a second facial identity based on the first texture map and the first position map, wherein the second facial identity is different from the first facial identity.

Abstract: A system includes a machine learning (ML) model-based video encoder configured to receive an uncompressed video sequence including multiple video frames, determine, from among the multiple video frames, a first video frame subset and a second video frame subset, encode the first video frame subset to produce a first compressed video frame subset, and identify a first decompression data for the first compressed video frame subset. The ML model-based video encoder is further configured to encode the second video frame subset to produce a second compressed video frame subset, and identify a second decompression data for the second compressed video frame subset. The first decompression data is specific to decoding the first compressed video frame subset but not the second compressed video frame subset, and the second decompression data is specific to decoding the second compressed video frame subset but not the first compressed video frame subset.

Abstract: According to one implementation, an image enhancement system includes a computing platform including a hardware processor and a system memory storing a software code configured to provide a normalizing flow based generative model trained using an objective function. The hardware processor executes the software code to receive an input image, transform the input image to a latent space representation of the input image using the normalizing flow based generative model, and perform an optimization of the latent space representation of the input image to identify an enhanced latent space representation of the input image. The software code then uses the normalizing flow based generative model to reverse transform the enhanced latent space representation of the input image to an enhanced image corresponding to the input image.

Abstract: A system includes a computing platform having a hardware processor and a memory storing a software code and a neural network (NN) having multiple layers including a last activation layer and a loss layer. The hardware processor executes the software code to identify different combinations of layers for testing the NN, each combination including candidate function(s) for the last activation layer and candidate function(s) for the loss layer. For each different combination, the software code configures the NN based on the combination, inputs, into the configured NN, a training dataset including multiple data objects, receives, from the configured NN, a classification of the data objects, and generates a performance assessment for the combination based on the classification. The software code determines a preferred combination of layers for the NN including selected candidate functions for the last activation layer and the loss layer, based on a comparison of the performance assessments.

Abstract: The invention relates to a method for reconstructing an output image and/or a sequence of output images from raw image data, wherein, in at least one reconstruction chain between a recording of raw image data by means of an image recording device and a display of the output image and/or the sequence of output images on a display unit, at least one reconstruction step for reconstructing the output image and/or the sequence of output images is carried out, and wherein the result of the at least one reconstruction step is discarded when the result is recognized as invalid with respect to the fulfillment of a verification criterion.

Abstract: A system processing hard e executes a machine learning (ML) model-based video compression encoder to receive uncompressed video content and corresponding motion compensated video content, compare the uncompressed and motion compensated video content to identify an image space residual, transform the image space residual to a latent space representation of the uncompressed video content, and transform, using a trained image compression ML model, the motion compensated video content to a latent space representation of the motion compensated video content.

Abstract: The present invention lies in the field of simulation systems for medical treatments, in particular for treatments carried out on dialysis machines. To do so, a method with which medical treatments can be performed and evaluated without the involvement of real human patients is proposed.

Abstract: An image processing system includes a computing platform having processing hardware, a display, and a system memory storing a software code. The processing hardware is configured to execute the software code to receive a three-dimensional (3D) digital model, surround the 3D digital model with multiple virtual cameras oriented toward the 3D digital model, and generate, using the virtual cameras, a multiple renders of the 3D digital model. The processing hardware is further configured to execute the software code to generate a UV texture coordinate space for a surface projection of the 3D digital model, and to transfer, using the multiple renders, lighting color values for each of multiple surface portions of the 3D digital model to the UV texture coordinate space.

Abstract: A system includes a computing platform having processing hardware, and a system memory storing software code and one or more machine learning (ML) model(s) trained using contrastive learning based on a similarity metric. The processing hardware is configured to execute the software code to receive input data including a plurality of content segments, map, using the ML model(s), each of the plurality of content segments to a respective embedding in a continuous vector space to provide a plurality of mapped embeddings, and perform one of a classification or a regression of the content segments using the plurality of mapped embeddings. The processing hardware is also configured to execute the software code to discover, based on the classification or the regression, at least one new label for characterizing the plurality of content segments.

Abstract: The present invention relates to door modules comprising at least one single-layer, continuous fiber-reinforced semifinished fiber matrix product which is obtained by impregnating semifinished fiber products with at least one thermoplastic, and to the use thereof in vehicles or in buildings.

Abstract: Absorbent laminates and folded multi-layer absorbent cores including one or more of the present absorbent laminates The present absorbent laminates comprise an absorbent layer between two laminate layers, at least one of which absorbent laminates including a spunlace nonwoven. Some of the present multi-layer absorbent cores are folded to define a channel running longitudinally along the core to enhance liquid distribution and absorption.

Abstract: A milling roll of a roll mill may have a roll main body and an edge element that is attached to an end region of a roll end of the milling roll and extends in a radial direction beyond a surface of the roll main body. The edge element comprises a base flange and a spacer element that is attached thereto. Wear protection elements can be attached to the spacer element. Furthermore, a roll mill may have a first milling roll and a second milling roll that are arranged opposite one another and can be driven in opposite directions. A milling gap is disposed between the milling rolls, and an edge element on one of the milling rolls may extend over the milling gap and at least partially cover an end side of the opposite milling roll.

Abstract: A roller mill for comminuting material may include first and second grinding rollers disposed opposite one another and drivable in opposite directions and a milling gap between the grinding rollers. An end region of at least the first grinding roller comprises a peripheral element that extends across the milling gap and at least partially covers an end side of the second grinding roller. A scraping element for at least partially removing material is disposed on the end region of the first grinding roller. A measuring device can determine a layer thickness of ground material on at least one of the grinding rollers and/or determine stress in the peripheral element. A control device connected to the measuring device is configured, based on the determined layer thickness and/or stress, to decrease or increase an amount of space between the scraping element and one of the grinding rollers.

Abstract: A method for generating a brightness contribution for a picture element of an image includes providing a first data record including data which describe the effect on light rays of the lens, providing a second data record including data about a point of incidence of a light ray on the image recorder and about a virtual front surface, providing a transformation rule, calculating a first point of intersection of the light ray with the virtual front plane and a direction of the light ray at the first point of intersection by applying the transformation rule to the first and second data records, determining the brightness contribution of the light ray, storing an information item regarding the calculated brightness contribution. The first data record includes data about a second surface and the second data record includes data about a second point of intersection of the light beam with the second surface.

Abstract: A gray zone prevention circuit includes: a first gain stage circuit including a first input terminal and a first output terminal, the first gain stage circuit amplifies a feedback signal received at the first input terminal and generates an amplified signal at the first output terminal; a second gain stage circuit including a terminal that is coupled to the first output terminal for receiving the amplified signal and a second output terminal, where the second gain stage circuit is configured to generate a monitored signal based on the amplified signal; a feedback circuit coupled between the second output terminal and the first input terminal and configured to convert the monitored signal into the feedback signal; and a comparator circuit including a monitoring node coupled to the first output terminal for receiving the amplified signal, wherein the comparator circuit is configured to monitor the monitored signal indirectly via the amplified signal.