Abstract: A system and a method for training a model to be used for semantic segmentation of images, comprising: a—obtaining (S01) a first plurality of foggy images (101), b—training (S02) a classification model for estimating fog density, c—classifying (S03) a second plurality of images (101) having light fog, d—obtaining (S04) a third plurality of foggy images (103) having light fog, e—training (S05) a semantic segmentation model using the third plurality of foggy images, f—applying (S06) the semantic segmentation model to the second plurality of foggy (102) images to obtain semantic segmentations (102?), g—obtaining (S07) a fourth plurality of foggy images (104) having dense fog, h—training (S08) using the previously obtained foggy images.

Abstract: The present invention relates to a dinitroxide biradical compound of general formula (I): In formula (I), A is a carbon, an ammonium, or a phosphonium; each of A1 to A6 is a single bond, O, N, N(O), S, S(O), SO2, C(O), or a (C1-C4) alkyl chain; and Z1 and Z2 are selected from R1 and R2 in combination, so that there is always at least one R2 group that is substituted with R3 group. In Z1 and Z2, R1 is an H, an aryl, or a heteroaryl; R2 is an alkyl chain, an alkenyl chain, an alkynyl chain, a cycloalkyl, a heterocycloalkyl, an aryl, or a heteroaryl; and R3 is an alkyl chain, an alkenyl chain, an alkynyl chain, a cycloalkyl, a heterocycloalkyl, an aryl, a heteroaryl, an ether, an ester, or an azide. When Z1 and Z2 are combined together with the same carbon atom of the nitroxide ring to which they are bonded, they form a spirocycloalkyl or a spiroheterocycloalkyl, the spirocycloalkyl or the spiroheterocycloalkyl being substituted with R3.

Abstract: The invention relates to a process for assembling DNA parts into multi-kilo base long synthetic DNA constructs. The process generates multiple, synonymous DNA parts in parallel and selects in a combinatorial assembly approach for those sequence variants with the best synthesis and assembly feasibility. DNA parts are sequence optimized and partitioned into synonymous variant designs that serve as redundant building units for higher order DNA assembly. The major stages of the process are: computational partitioning and synonymous recoding of the DNA design, DNA synthesis of sequence variants pools, serial PGR to isolate sets of DNA parts and higher order assembly. As the higher-order assembly does no longer depends on successful synthesis of each DNA part, large-scale DNA designs can be quickly completed allowing for cost-effective and highly parallelised assembly of synthetic bio-designs.

Abstract: The present invention relates to a method for providing an embedded mammalian cell, comprising the steps of providing an alginate sulfate in aqueous solution; reacting the alginate sulfate to form a hydrogel in a gelation step, providing a precursor cell, and embedding the precursor cell in the sulfated alginate hydrogel in an embedding step, thus yielding an sulfated alginate hydrogel embedded cell. The invention further relates to sulfated alginate hydrogels, and cellular grafts comprising a mammalian cell embedded in sulfated alginate hydrogel.

Abstract: Techniques are disclosed for image matting. In particular, embodiments decompose the matting problem of estimating foreground opacity into the targeted subproblems of estimating a background using a first trained neural network, estimating a foreground using a second neural network and the estimated background as one of the inputs into the second neural network, and estimating an alpha matte using a third neural network and the estimated background and foreground as two of the inputs into the third neural network. Such a decomposition is in contrast to traditional sampling-based matting approaches that estimated foreground and background color pairs together directly for each pixel. By decomposing the matting problem into subproblems that are easier for a neural network to learn compared to traditional data-driven techniques for image matting, embodiments disclosed herein can produce better opacity estimates than such data-driven techniques as well as sampling-based and affinity-based matting approaches.

Abstract: The invention relates to a method for the manufacturing of a carrying device, a carrying device, a system for detection and a method for detection of at least one physical parameter and/or chemical composition. A method for the manufacturing of a carrying device for reception of at least one sensor is presented, in which a receiving body with a surface to be coated is provided. In the receiving body, a space is provided which is open on the side of this surface. A second body with a sealing surface is provided. This sealing surface is positioned in such a manner that the surface to be coated of the receiving body is sealed by the second body at least in the circumference of the space in the receiving body. A formable or moldable filling material is provided in the space in such a manner that the filling material forms a surface shaped complementary to the sealing surface of the second body, closing the space.

Abstract: According to one implementation, a pose estimation and body tracking system includes a computing platform having a hardware processor and a system memory storing a software code including a tracking module trained to track motions. The software code receives a series of images of motion by a subject, and for each image, uses the tracking module to determine locations corresponding respectively to two-dimensional (2D) skeletal landmarks of the subject based on constraints imposed by features of a hierarchical skeleton model intersecting at each 2D skeletal landmark. The software code further uses the tracking module to infer joint angles of the subject based on the locations and determine a three-dimensional (3D) pose of the subject based on the locations and the joint angles, resulting in a series of 3D poses. The software code outputs a tracking image corresponding to the motion by the subject based on the series of 3D poses.

Abstract: The device provides a resistive memory device for storing elements of hyper-dimensional vectors, in particular digital hyper-dimensional, as conductive statuses in components in particular in 2D-memristors, of the resistive memory device, wherein the resistive memory device provides a first crossbar array of the components, wherein the components are memristive 2D components addressable by word-lines and bit-lines, and a peripheral circuit connected to the word-lines and bit-lines and adapted for encoding operations by activating the word-lines and bit-lines sequentially in a predefined manner.

Abstract: The invention relates to a device (1), a method, and a kit for analysing liquid samples. The device (1) comprises a sample layer (111) having a plurality of liquid permeable test sites (112) separated by a liquid impermeable barrier region (113), and an inlet part (2) comprising a plurality of inlet channels (211), which lead to respective test sites (112) of the sample layer (111), such that a flow connection between said inlet channels (211) and said respective test sites (112) is established or can be established, wherein said inlet channels (211) comprise first openings (218) and second openings (219), wherein a second surface area defined by the positions of said second openings (219) is smaller than a first surface area defined by the positions of said first openings (218) The invention further relates to a method for functionalizing a sample layer (111).

Abstract: The invention relates to a method, a device, and a system for manipulating portions of a rigid body (1), and a computer program for the execution of the steps of the method for manipulating portions (20) of a rigid body (1). The method comprises the steps of providing a rigid body (1) comprising ultrathin-sectionable material, cutting an ultrathin portion (20) from the rigid body (1), providing the portion (20) at the surface of a liquid, providing at least one magnetic device (4), which produces a magnetic field or which is able to produce a magnetic field, using the magnetic field in order to provide a magnetic force, and using the magnetic force to move the portion (20) at the surface of the liquid. The device, the system and the computer program are adapted to carry out the method for manipulating portions (20) of a rigid body (1).

Abstract: A method for analysing an analyte (3) using surface enhanced Raman spectroscopy (SERS), comprising the following steps: (a) providing an essentially flat or topologically structured metal surface (1) of a SERS-active metal; (b) depositing the analyte (3) or an open pore matrix material (5) on the surface (1); (c) depositing a multitude of nano-droplets (2) of a SERS-active metal on top of the analyte (3) or the open pore matrix material (5), respectively; and (d) spectroscopically analysing, by scanning laser irradiation and using SERS, the analyte sandwiched between the surface (1) and the multitude of nano-droplets (2). The diameter of the nano-droplets (2) is in the range of 5-70 nm, and the distance between adjacent nano-droplets (2) is smaller than their diameter, and wherein step c) is carried out by PVD or by sputtering SERS-active metal.

Abstract: A mobile manufacturing device and a method for producing three-dimensional lattice or mesh structures, wherein the mobile manufacturing device is a mobile device adapted to move along a three-dimensional structure to be produced, in particular while performing both translation and rotation movements in three-dimensional space, and wherein the mobile manufacturing device includes a means for detection of its position relative to the dimensional structure to be produced, so that the mobile manufacturing device is able to produce the three-dimensional structure as an autonomous robot.

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 method of printing a print pattern onto a substrate with a print head comprises a plurality of nozzles, where the print head has a rectangular active print head area which includes all of the nozzles. The active print head area is delimited by four sides defining a primary and a secondary direction. The method comprises i) decomposing the print pattern into a plurality of print pattern segments that have dimensions along the primary and secondary direction which are smaller than the dimensions of the active print head area along the primary and secondary direction; ii) assigning each print pattern segment to exactly one nozzle; iii) causing each nozzle to print the print pattern segment assigned to said nozzle. The print head is moved during printing of each print pattern segment within an area that is smaller than said active print head area.

Abstract: A chemical inhibitor is used for reducing and/or preventing swelling of anhydrite containing rock material, wherein, the chemical inhibitor is used during construction work, especially in tunnel construction through rock material, especially in drill and blast tunnel construction work and/or in tunnel construction with a tunnel boring machine.

Abstract: Techniques are disclosed for estimating poses from images. In one embodiment, a machine learning model, referred to herein as the “detector,” is trained to estimate animal poses from images in a bottom-up fashion. In particular, the detector may be trained using rendered images depicting animal body parts scattered over realistic backgrounds, as opposed to renderings of full animal bodies. In order to make appearances of the rendered body parts more realistic so that the detector can be trained to estimate poses from images of real animals, the body parts may be rendered using textures that are determined from a translation of rendered images of the animal into corresponding images with more realistic textures via adversarial learning. Three-dimensional poses may also be inferred from estimated joint locations using, e.g., inverse kinematics.

Abstract: A device for detecting and evaluating movements of a user comprises a processing unit, a data memory, a sensor unit with at least one biometric sensor, and a fastening device for fastening the sensor unit on an arm of the user. The processing unit is configured to detect movement data from movement signals generated by the sensor unit and to store the movement data in the data memory. The device further comprises an evaluation module with which secondary movement data, which represent passive arm movements of the user, are filterable out of the movement data. The device enables the detection and evaluation of movements of an arm of the user in an efficient way.

Abstract: The invention relates to a polymer structure (1) formed by at least a polymer, wherein said structure (1) comprises a volume (2) and a surface (3), wherein said polymer comprises a plurality of polymer chains connected by linkings, characterized by a linking density, wherein said linking density increases, particularly monotonously, from the surface (3) into the volume (2) of the polymer structure (1).

Abstract: The present invention is notably directed to a computerized method for providing real-time feedback to a user from states of a model physical system, or MPS, via a computerized system comprising one or more processors and a user interface system, or UIS. The method comprises the following steps, each performed via the one or more processors. Configuration inputs are repeatedly received, to modify a configuration of the MPS, said inputs including user inputs received via said UIS. While receiving said configuration inputs: configurations of the MPS are updated based on the configuration inputs received; and a state of the MPS is repeatedly computed, whereby each computed state corresponds to a latest updated configuration that was available before starting to compute said each computed state.

Abstract: Systems and methods for applying a pulsed electromagnetic fields. A system for applying a pulsed electromagnetic field (PEMF) to a cell, comprising: a sensor for obtaining a characteristic of the cell; a memory module having stored therein a plurality of characteristics and PEMF efficacy window data, each characteristic having its corresponding PEMF efficacy window data; a pulse generator coupled to a set of PEMF coils and configured to generate an output of electrical pulses to drive the set of PEMF coils; and a controller in communication with the sensor, the memory module and the pulse generator, wherein the controller is configured to: retrieve, from the memory module, the PEMF efficacy window data that corresponds to the characteristic of the cell obtained by the sensor; and control the output of the pulse generator based on the retrieved PEMF efficacy window data such that the set of PEMF coils apply a PEMF in accordance with the PEMF efficacy window data.