Abstract: A method of generating a training data set for training an image matting machine learning model includes receiving a plurality of foreground images, generating a plurality of composited foreground images by compositing randomly selected foreground images from the plurality of foreground images, and generating a plurality of training images by compositing each composited foreground image with a randomly selected background image. The training data set includes the plurality of training images.

Abstract: The present disclosure relates to a catalyst composition comprising copper and iron on a support for use in a process for the synthesis of higher alcohols from a syngas feed stream comprising hydrogen and carbon monoxide, the catalyst composition being remarkable in that the support is one or more zeolite, in that the total content of iron and copper is ranging from 1 to 10 wt. % based on the total weight of the catalyst composition and as determined by inductively coupled plasma optical emission spectroscopy, in that the Cu/Fe bulk molar ratio is ranging from 1.1:1.0 to 5.0:1.0 as determined by XRF spectroscopy.

Abstract: The invention relates to the manufacturing of porous polymer membranes by (a) providing pellets comprising a polymer matrix and particles in the ratio 90:10 to 10:90, (b) converting said pellets into a non-porous film by a solvent-free process; (c) removing said particles from said film with an aqueous composition to thereby obtain said membrane. The invention further relates to pellets useful in such manufacturing process as well as porous polymer membranes obtainable or obtained by such manufacturing process as well as textile materials and articles containing such membranes; to the use of such pellets, membranes, and articles.

Abstract: A device for producing a target material from starting material comprises a chamber, at least one trough, at least a first heating element being configured to heat the chamber such that starting material is vaporized, and at least one collecting vessel being configured to receive a condensate that will constitute the target material. The device optionally comprises at least a first source of negative pressure or at least a first supply device being in connection with the chamber being configured to evacuate the chamber or to supply an inert gas to the chamber. The device further comprises at least one condensation device, wherein said condensation device is configured to condensate the vaporized starting material, whereby the condensate is formed, and/or at least a first gate device being in connection with the chamber such, that the starting material is introducible into the chamber via said first gate device.

Abstract: An image processing system includes a computing platform having a hardware processor and a system memory storing an image augmentation software code, a three-dimensional (3D) shapes library, and/or a 3D poses library. The image processing system also includes a two-dimensional (2D) pose estimation module communicatively coupled to the image augmentation software code. The hardware processor executes the image augmentation software code to provide an image to the 2D pose estimation module and to receive a 2D pose data generated by the 2D pose estimation module based on the image. The image augmentation software code identifies a 3D shape and/or a 3D pose corresponding to the image using an optimization algorithm applied to the 2D pose data and one or both of the 3D poses library and the 3D shapes library, and may output the 3D shape and/or 3D pose to render an augmented image on a display.

Abstract: A system includes a computing platform having a hardware processor, and a system memory storing a software code and a content labeling predictive model. The hardware processor is configured to execute the software code to scan a database to identify content assets stored in the database, parse metadata stored in the database to identify labels associated with the content assets, and generate a graph by creating multiple first links linking each of the content assets to its corresponding label or labels. The hardware processor is configured to further execute the software code to train, using the graph, the content labeling predictive model, to identify, using the trained content labeling predictive model, multiple second links among the content assets and the labels, and to annotate the content assets based on the second links.

Abstract: A computer-implemented method (200) of radiation events localizations is indicated for a pixelated radiation detector (10) having a scintillator array (24) of scintillator array elements (26) arranged in an (m)×(n) array, and an optical sensor array (28) of optical sensors (30) arranged in a (q)×(z) array and coupled to the scintillator array (24) in light sharing mode. The method includes the steps of sampling (72) spatial intensity distributions of scintillation photons emitted by the scintillator array (24) in response to multiple incident radiation events; performing a clustering analysis (76) based on the sampled spatial intensity distributions, to obtain clusters (84) of radiation events attributed to scintillator array elements (26), wherein the dimension of the sampled spatial intensity distributions correspond to the (q)×(z) dimensions of the optical sensor array (28), and determining the localization of the radiation events based on the clustering analysis (76).

Abstract: According to a first aspect of the invention, there is provided a method for operating a multicopter experiencing a failure during flight, the multicopter comprising a body, and at least four effectors attached to the body, each operable to produce both a torque and a thrust force which can cause the multicopter to fly when not experiencing said failure.

Abstract: A modeling engine generates a prediction model that quantifies and predicts secondary dynamics associated with the face of a performer enacting a performance. The modeling engine generates a set of geometric representations that represents the face of the performer enacting different facial expressions under a range of loading conditions. For a given facial expression and specific loading condition, the modeling engine trains a Machine Learning model to predict how soft tissue regions of the face of the performer change in response to external forces applied to the performer during the performance. The modeling engine combines different expression models associated with different facial expressions to generate a prediction model. The prediction model can be used to predict and remove secondary dynamics from a given geometric representation of a performance or to generate and add secondary dynamics to a given geometric representation of a performance.

Abstract: A neural network, a system using this neural network and a method for training a neural network to output a description of the environment in the vicinity of at least one sound acquisition device on the basis of an audio signal acquired by the sound acquisition device, the method including: obtaining audio and image training signals of a scene showing an environment with objects generating sounds, obtaining a target description of the environment seen on the image training signal, inputting the audio training signal to the neural network so that the neural network outputs a training description of the environment, and comparing the target description of the environment with the training description of the environment.

Abstract: A method for producing a nano-porous membrane with one or up to four graphene layers, pores in the membrane having an average pore size in the range of 0.2-50 or 0.3-10 nm, wherein the method involves the following steps: a) generation of a contiguous, essentially non-porous membrane with one or up to four graphene layers; b) distributed point wise defect creation in the non-porous membrane with one or up to four graphene layers by way of irradiation; c) generation and successive growth of the pores at the defects generated in step b) by thermal annealing in the gas phase, e.g. under 02 at a temperature in the range of 250° C. to less than 400° C.

Abstract: A system for hyper-dimensional computing for inference tasks may be provided. The device comprises an item memory for storing hyper-dimensional item vectors, a query transformation unit connected to the item memory, the query transformation unit being adapted for forming a hyper-dimensional query vector from a query input and hyper-dimensional base vectors stored in the item memory, and an associative memory adapted for storing a plurality of hyper-dimensional profile vectors and for determining a distance between the hyper-dimensional query vector and the plurality of hyper-dimensional profile vectors, wherein the item memory and the associative memory are adapted for in-memory computing using memristive devices.

Abstract: Techniques are disclosed for changing the identities of faces in images. In embodiments, a tunable model for changing facial identities in images includes an encoder, a decoder, and dense layers that generate either adaptive instance normalization (AdaIN) coefficients that control the operation of convolution layers in the decoder or the values of weights within such convolution layers, allowing the model to change the identity of a face in an image based on a user selection. A separate set of dense layers may be trained to generate AdaIN coefficients for each of a number of facial identities, and the AdaIN coefficients output by different sets of dense layers can be combined to interpolate between facial identities. Alternatively, a single set of dense layers may be trained to take as input an identity vector and output AdaIN coefficients or values of weighs within convolution layers of the decoder.

Abstract: One embodiment of the present invention sets forth a technique for performing deinterlacing. The technique includes separating a first interlaced video frame into a first sequence of fields ordered by time, the first sequence of fields including a first field. The technique also includes generating, by applying a deinterlacing network to a first field in the first sequence, a second field that is missing from the first sequence of fields and is complementary to the first field. The technique further includes constructing a progressive video frame based on the first field and the second field.

Abstract: Disclosed is a thermochemical reactor system and method for a temperature swing cyclic process with integrated heat recovery having at least two modules, wherein each module includes at least one chemical reaction zone and at least one thermal energy storage unit. The at least two modules are operationally connected for at least one heat transfer fluid for transporting heat between the two modules. Each chemical reaction zone includes at least one reacting material that undergoes in a reversible manner an endothermic reaction at temperature Tendo and an exothermic reaction at temperature Texo, wherein Tendo and Texo differ from each other. The at least one reacting material is provided in at least one encapsulation within each of the chemical reaction zones such that a contact of the reacting material and the at least one heat transfer fluid is avoided.

Abstract: In various embodiments, a relevance application quantifies how a trained neural network operates. In operation, the relevance application generates a set of input distributions based on a set of input points associated with the trained neural network. Each input distribution is characterized by a mean and a variance associated with a different neuron included in the trained neural network. The relevance application propagates the set of input distributions through a probabilistic neural network to generate at least a first output distribution. The probabilistic neural network is derived from at least a portion of the trained neural network. Based on the first output distribution, the relevance application computes a contribution of a first input point included in the set of input points to a difference between a first output point associated with a first output of the trained neural network and an estimated mean prediction associated with the first output.

Abstract: A sensor assembly adapted to be embedded in a concrete structure comprising a body, and at least one sensor configured to determine parameters related to the durability of a concrete structure, wherein the sensor is arranged at least partially within the body, wherein the body comprises a shell covering the outer surface of the body and consisting of or comprising a mineral material, and wherein the sensor assembly has a rounded shape, particularly an ellipsoid or spherical shape, and a first extension extending along a first axis being 90 mm or less. A concrete structure is also provided having at least one sensor assembly.

Abstract: In a method for aromatizing an alicyclic region of a cannabinoid, especially in enantiopure, scalemic and/or racemic form, in particular for aromatizing the cyclohexene group in ?9-THC-C5, ?9-THCA-C5 A, ?9-THCV-C3, ?9-THCVA-C5 A, and/or scalemic or racemic mixtures of these substances, wherein an oxidizing agent is added to the cannabinoid, sulfur is used as the oxidizing agent.

Abstract: The invention is notably directed to a method for encoding information. This method first comprises generating an encryption key according to polymorphic features of nucleic acids from one or more entities. Next, information is encrypted based on the generated key. Finally, the encrypted information is encoded into synthetic DNA. Another aspect concerns a method for retrieving information. Consistently with the above encoding scheme, synthetic DNA in provided, which encodes encrypted information. Such information is read by sequencing the synthetic DNA and by decrypting the information read using a decryption key. The latter is generated according to polymorphic features of nucleic acids from one or more entities (e.g., from the legitimate individual(s) requesting access to information). Thus, the encoded information cannot be interpreted unless a suitable decryption key is available. The invention is further directed to related DNA samples and systems, including DNA vaults.

Abstract: The invention relates to a rotational shear rheometer (1) comprising a first plate (10), a first flexure (11) pivotally connecting the first plate (10) to a support (60), a piezoelectric element (40) extending along a longitudinal axis (L) arranged tangentially in respect of the first plate (10), wherein said piezoelectric element (40) is configured to elongate and/or undergo compression along said longitudinal axis (L), and wherein said first end (41) is mechanically coupled to said first plate (10), wherein said first flexure (11) comprises a first flexural element (12) and a second flexural element (13) non-parallel to the first flexural element (12), wherein the first flexural element (12) and the second flexural element (13) connect said first plate (10) to said support (60), extend radially in respect of the pivot axis (P) and intersect with the pivot axis (P), wherein said piezoelectric element (40) is further configured as a sensor capable of detecting a torque acting on said first plate (10).