Abstract: Methods, system and computer program product, the method comprising: from high level language code (HLLC), receiving a request for reading a data set from a tape onto an object storage connected over TCP/IP to a mainframe; from the HLLC, allocating a data set on a tape comprising information to be imported, the allocation being in a format of the stored data set record and associated with a JFCB, the tape is mounted in SL mode; updating the JFCB to BLP mode; reading from the tape VOL1 data, and for each stored file initiating by the HLLC: reading HDR1/2, content block-by-block; EOF1/2 of the file; organizing the VOL1, HDR1, HDR2, content, EOF1 and EOF2 in the object storage; and closing the tape, wherein said reading is performed without setting a JES of the mainframe to BLP mode, and said reading is performed without unmounting the tape after each file.

Abstract: Systems, methods, and non-transitory computer readable media including instructions for presenting location-based content are described. Presenting location-based content includes obtaining an indication of a current physical location of an extended reality appliance; providing the indication to a first server that maps physical locations to a plurality of content addresses; receiving from the first server, at least one specific content address associated with the current physical location; using the at least one specific content address to access a second server; receiving content, associated with the current physical location, from the second server; and presenting the content via the extended reality appliance, while the extended reality appliance is in the current physical location.

Abstract: A system and a method for content localization in moving vehicles may include receiving from within a moving vehicle first acceleration data captured using a first sensor included in an extended reality appliance mountable on a head of a wearer. The first acceleration data includes a first component associated with movement of the head of the wearer with respect to the vehicle and a second component associated with movement of the vehicle. The system also includes receiving from within the moving vehicle second acceleration data captured using a second sensor included in a personal input device. The personal input device may be a non-vehicle component configured to be paired with the extended reality appliance. The system may segregate the second component from the first component using the first acceleration data and the second acceleration data to thereby isolate the head acceleration with respect to the vehicle from the vehicle acceleration.

Abstract: An object detection system may generate regions of interest (ROIs) from an input image that can be processed by a wide range of object detectors. According to the techniques described herein, an image is processed by a light-weight neural network (e.g., a heatmap network) that outputs object center and object scale heat-maps. The heatmaps are processed to define ROIs that are likely to include objects. Overlapping ROIs are then merged to reduce the aggregate size of the ROIs, and merged ROIs are downscaled to a reduced set of pre-defined resolutions. Fully-convolutional, high-accuracy object detectors may then operate on the downscaled ROIs to output accurate detections at a fraction of the computations by operating on a reduced image. For example, fully-convolutional, high-accuracy object detectors may operate on a subset of the entire image (e.g., cropped images based on ROIs) thus reducing computations otherwise performed over the entire image.

Abstract: An object detection system may generate regions of interest (ROIs) from an input image that can be processed by a wide range of object detectors. According to the techniques described herein, an image is processed by a light-weight neural network (e.g., a heatmap network) that outputs object center and object scale heat-maps. The heatmaps are processed to define ROIs that are likely to include objects. Overlapping ROIs are then merged to reduce the aggregate size of the ROIs, and merged ROIs are downscaled to a reduced set of pre-defined resolutions. Fully-convolutional, high-accuracy object detectors may then operate on the downscaled ROIs to output accurate detections at a fraction of the computations by operating on a reduced image. For example, fully-convolutional, high-accuracy object detectors may operate on a subset of the entire image (e.g., cropped images based on ROIs) thus reducing computations otherwise performed over the entire image.

Abstract: A system of convolutional neural networks (CNNs) that synthesize middle non-existing frames from pairs of input frames includes a coarse CNN that receives a pair of images acquired at consecutive points of time, a registration module, a refinement CNN, an adder, and a motion-compensated frame interpolation (MC-FI) module. The coarse CNN outputs from the pair of images a previous feature map, a next feature map, a coarse interpolated motion vector field (IMVF) and an occlusion map, the registration module uses the coarse IMVF to warp the previous and next feature maps to be aligned with pixel locations of the IMVF frame, and outputs registered previous and next feature maps, the refinement CNN uses the registered previous and next feature maps to correct the coarse IMVF, and the adder sums the coarse IMVF with the correction and outputs a final IMVF.

Abstract: A system of convolutional neural networks (CNNs) that synthesize middle non-existing frames from pairs of input frames includes a coarse CNN that receives a pair of images acquired at consecutive points of time, a registration module, a refinement CNN, an adder, and a motion-compensated frame interpolation (MC-FI) module. The coarse CNN outputs from the pair of images a previous feature map, a next feature map, a coarse interpolated motion vector field (IMVF) and an occlusion map, the registration module uses the coarse IMVF to warp the previous and next feature maps to be aligned with pixel locations of the IMVF frame, and outputs registered previous and next feature maps, the refinement CNN uses the registered previous and next feature maps to correct the coarse IMVF, and the adder sums the coarse IMVF with the correction and outputs a final IMVF.

Abstract: A method of matching stereo images includes: determining, first information indicating which orientations are dominant in each pixel location of a right image and a left image among the stereo images; determining second information indicating for each pixel location in the right image and the left image whether it lies on a pure horizontal edge; detecting points in the left image and the right image based on the first information and the second information; and performing a matching on the left image and the right image using the detected points to estimate a sparse disparity map.

Abstract: A method of matching stereo images includes: determining, first information indicating which orientations are dominant in each pixel location of a right image and a left image among the stereo images; determining second information indicating for each pixel location in the right image and the left image whether it lies on a pure horizontal edge; detecting points in the left image and the right image based on the first information and the second information; and performing a matching on the left image and the right image using the detected points to estimate a sparse disparity map.