Abstract: Methods are disclosed for generating a training dataset of concealed shapes and corresponding unveiled shapes of a body for training a neural network. These methods may include generating with the aid of computing means a first dataset comprising a plurality of first surface representations representative of a plurality of bare shapes of a plurality of bodies. The plurality of bare shapes are concealed virtually by means of a computer implemented program in order to obtain a plurality of simulated concealed shapes of the plurality of bodies. The plurality of simulated concealed shapes are applied to a scanning simulator, the scanning simulator generating a second dataset comprising a plurality of second surface representations representative of the plurality of simulated concealed shapes.

Abstract: Method of determining a position (f a sensor device movable relative to a magnetic source, or vice versa; the sensor device comprising at least two magnetic sensors; the method comprising the steps of: a) obtaining a plurality of magnetic sensor signals from said magnetic sensors; b) determining or estimating the position of the sensor device based on said plurality of sensor signals or signals derived therefrom; wherein step b) comprises: determining said position (sing an artificial neural network; the artificial neural network being a recurrent neural network trained for determining said position using at most three hundred trainable parameters per degree of freedom. A position sensor system. A position sensor device.

Abstract: A contour-cutting machine for cutting a foam block, such as a polyurethane foam block into foam rows, such as pillows, has high productivity. The contour-cutting machine has multiple cutting tools that can be operated simultaneously to cut the foam block. Each cutting tool has a blade that cuts through the foam block. For example, an upper cutting tool and a lower cutting tool mounted over a pillar one above each other can be positioned at a predefined distance from each other and locked together. Each cycle of the process makes two cuts with the two cutting tools in the foam block.

Abstract: Methods are disclosed for generating a training dataset of concealed shapes and corresponding unveiled shapes of a body for training a neural network. These methods may include generating with the aid of computing means a first dataset comprising a plurality of first surface representations representative of a plurality of bare shapes of a plurality of bodies. The plurality of bare shapes are concealed virtually by means of a computer implemented program in order to obtain a plurality of simulated concealed shapes of the plurality of bodies. The plurality of simulated concealed shapes are applied to a scanning simulator, the scanning simulator generating a second dataset comprising a plurality of second surface representations representative of the plurality of simulated concealed shapes.

Abstract: A hash-based distributed video coding architecture has an encoder with an input video sequence organized in Groups of Pictures (GOPs) decomposed into key frames using H264/AVC Intra frame coding. The Wyner-Ziv (WZ) frames are encoded in two parts, a hash layer and a WZ layer. The WZ frames are quantized and each and then decorrelated using spatio-temporal prediction and entropy coded. The intra bit stream is H264/AVC decoded and the intra frames are stored in a reference frame buffer. The hash is decoded by inverting the tasks applied at the encoder, and the obtained bit planes are stored. Overlapped Block Motion Estimation and Probalistic Compensation (OBMEPC) is used to estimate the missing bit planes in the side-information. The decoder utilizes the hash information and the side information frame created by OBMEPC to perform online estimation of the correlation channel, and produces soft estimates used to decode the WZ bit planes.

Abstract: In the motion estimation and compensation process for video frames, blocks O of pixels are considered. A number k of bit planes in a block O in a video frame F are compared with blocks OR in reference frames (FR). The best matching block (ORM) is determined in the reference frames (FR). Subsequently, a weight value (Wxij) is calculated for the best matching block (ORM) based on the ratio of valid pixels therein. The residual pixel values (Vxij) extracted from the best matching block (ORM) and corresponding weight values (Wxij) are stored in a pixel prediction array (120). The pixel array is used for motion compensation of at least the luminance component of valid pixels. Invalid pixels are reconstructed from surrounding pixel values.

Abstract: A bit stream representing n-dimensional data structures may be encoded and decoded. A part of the data can be mappable within predefined similarity criteria to a part of the data of another data structure. The similarity criteria may include, a spatial or temporal shift of the data. The data structures are typically sequential video frames such as is used in motion estimation and/or compensation of moving pictures, and a part of the data structure may be a block of data within a frame. The shift may be any suitable shift such as linear translation, rotation, or change of size. Digital filtering may be applied to a reference or other frame of data to generate subbands of a set of subbands of an overcomplete representation of the frame by calculations performed at single rate. The digital filtering may be implemented in a separate filter module or in software.

Abstract: A method and a device for transmission and/or reception of digital signals are disclosed. In one embodiment, the method comprises quantizing a source digital signal to generate with different quantizations at least a first and a second bit-stream, of which at least one bit-stream is generated by an embedded quantization, transmitting at least one of the at least first and second bit-streams and generating a dequantized digital signal from at least parts of one of the transmitted at least first and second bit streams, whereby if in the generation of the dequantized digital signal the parts of the at least first and second bit-streams are combined, the combined dequantized signal is generated by an embedded dequantizer having at least two quantization levels and having at least one quantization interval at each quantization level which is finer than quantization intervals for dequantizing any of the at least first and second bit-streams.

Abstract: A method and apparatus is described for coding motion information in video processing of a stream of image frames and for avoiding the drift problem. The method or apparatus is for providing motion vectors of at least one image frame, and for coding the motion vectors to generate a quality-scalable representation of the motion vectors. The quality-scalable representation of motion vectors can comprise a set of base-layer motion vectors and a set of one or more enhancement-layers of motion vectors. The method of decoding and a decoder for such coded motion vectors as part of receiving and processing a bit stream at a receiver includes the base-layer of motion vectors being losslessly decoded, while the one or more enhancement layers of motion vectors are progressively received and decoded, optionally including progressive refinement of the motion vectors, eventually up to their lossless reconstruction.

Abstract: A method and a device for transmission and/or reception of digital signals are disclosed. In one embodiment, the method comprises quantizing a source digital signal to generate with different quantizations at least a first and a second bit-stream, of which at least one bit-stream is generated by an embedded quantization, transmitting at least one of the at least first and second bit-streams and generating a dequantized digital signal from at least parts of one of the transmitted at least first and second bit streams, whereby if in the generation of the dequantized digital signal the parts of the at least first and second bit-streams are combined, the combined dequantized signal is generated by an embedded dequantizer having at least two quantization levels and having at least one quantization interval at each quantization level which is finer than quantization intervals for dequantizing any of the at least first and second bit-streams.

Abstract: Method and apparatus for encoding and decoding a bit stream comprising a representation of a sequence of n-dimensional data structures or matrices, in which n is typically 2. A part of the data of one data structure of the sequence can be mappable within predefined similarity criteria to a part of the data of a another data structure of the sequence. The similarity criteria may include, for instance, a spatial or temporal shift of the data within an image of a video signal such as is used in motion estimation and/or compensation of moving pictures, e.g., video images as well as coders and encoders for coding such a bitstream, e.g., a video signal for use in motion compensation and/or estimation. The data structures are typically video frames and a part of the data structure may be a block of data within a frame.

Abstract: A method of determining the motion vector of a block of a video frame with respect to a reference video frame. The video frame and reference video frame comprising pixels, wherein each pixel has a pixel value. The method comprises first determining a plurality of sets of error norms, wherein each error norm within one of the sets is related to a different position of the block in the reference video frame, wherein the error norms are calculated by a norm of an error which is given by functions of the pixel values of the block of the video frame and the reference video frame, and wherein each set of the plurality of sets is related to a different function. Second, the motion vector is selected with the smallest error norm.