Abstract: An additive manufacturing system includes a print head, a first light source array, a second light source array, and a controller. The print head includes a first rotatable reflector and a second rotatable reflector. The first light source array is configured to direct multiple first light beams toward a build medium via the first rotatable reflector. The second light source array is configured to direct multiple second light beams toward the build medium via the second rotatable reflector. The controller is coupled to the print head, the first light source array, and the second light source array. The controller is configured to control activation of the first light beams and activation of the second light beams to selectively fuse portions of the build medium to form one or more objects from the build medium.

Abstract: Certain aspects of the present disclosure provide a method and apparatus for image sensing. Embodiments include a first plurality of light-sensing elements arranged in a first plurality of rows and a first plurality of columns in a first image sensing plane. Embodiments include a first lens element disposed within a position in a central row of the first plurality of rows and in a central column of the first plurality of columns. Embodiments include a second plurality of light-sensing elements arranged in a second plurality of rows in a second image sensing plane, different from the first image sensing plane. In certain embodiments, the first lens element is configured to redirect incident light reflected from a destination onto the second plurality of light-sensing elements in the second image sensing plane for use in guiding a steerable object toward the destination.

Abstract: An additive manufacturing system includes a print head, a first light source array, a second light source array, and a controller. The print head includes a first rotatable reflector and a second rotatable reflector. The first light source array is configured to direct multiple first light beams toward a build medium via the first rotatable reflector. The second light source array is configured to direct multiple second light beams toward the build medium via the second rotatable reflector. The controller is coupled to the print head, the first light source array, and the second light source array. The controller is configured to control activation of the first light beams and activation of the second light beams to selectively fuse portions of the build medium to form one or more objects from the build medium.

Abstract: Certain aspects of the present disclosure provide a method and apparatus for image sensing. Embodiments include a first plurality of light-sensing elements arranged in a first plurality of rows and a first plurality of columns in a first image sensing plane. Embodiments include a first lens element disposed within a position in a central row of the first plurality of rows and in a central column of the first plurality of columns. Embodiments include a second plurality of light-sensing elements arranged in a second plurality of rows in a second image sensing plane, different from the first image sensing plane. In certain embodiments, the first lens element is configured to redirect incident light reflected from a destination onto the second plurality of light-sensing elements in the second image sensing plane for use in guiding a steerable object toward the destination.

Abstract: A method for classifying at least one object of interest in a video is provided. The method includes accessing, using at least one processing device, a frame of the video, the frame including at least one object of interest to be classified, performing, using the at least one processing device, object detection on the frame to detect the object of interest, tracking, using the at least one processing device, the object of interest over a plurality of frames in the video over time using a persistent tracking capability, isolating, using the at least one processing device, a segment of the frame that includes the object of interest, classifying, using the at least one processing device, the object of interest by processing the segment using deep learning, and generating an output that indicates the classification of the object of interest.

Abstract: Systems, methods, and apparatus for converting serial video data for a graphics processing unit (GPU) interface are disclosed. In one or more embodiments, a disclosed method comprises receiving, by each of a plurality of gigabit multimedia serial link (GMSL) conversion modules, n-bit length serial video data from a plurality of high-resolution cameras respectively. The method further comprises converting, by each of the GMSL conversion modules, the n-bit length serial video data to m-bit length serial GMSL video data, where n is equal to twice m. Also, the method comprises receiving, by a GMSL to camera series interface (CSI) conversion unit, the m-bit length serial GMSL video data from each of the GMSL conversion modules. Further, the method comprises converting, with the GMSL to CSI conversion unit, the m-bit length serial GMSL video data to m-bit length serial CSI video data, which is compatible with the GPU interface.

Abstract: Systems, methods, and apparatus for converting serial video data for a graphics processing unit (GPU) interface are disclosed. In one or more embodiments, a disclosed method comprises receiving, by each of a plurality of gigabit multimedia serial link (GMSL) conversion modules, n-bit length serial video data from a plurality of high-resolution cameras respectively. The method further comprises converting, by each of the GMSL conversion modules, the n-bit length serial video data to m-bit length serial GMSL video data, where n is equal to twice m. Also, the method comprises receiving, by a GMSL to camera series interface (CSI) conversion unit, the m-bit length serial GMSL video data from each of the GMSL conversion modules. Further, the method comprises converting, with the GMSL to CSI conversion unit, the m-bit length serial GMSL video data to m-bit length serial CSI video data, which is compatible with the GPU interface.

Abstract: Systems and methods for generating radio frequency signals and/or microwave signals using a tunable optical source. An optical frequency comb including multiple optical components is received or generated based on an optical signal. A subset of optical components is selected from the multiple optical components. A detector array having two or more square law detectors is uniformly illuminated with the subset of optical components. Each square law detector of the detector array outputs an electrical signal having a difference frequency corresponding to a difference in frequency between the subset of optical components. A radio frequency or microwave output signal with a narrow bandwidth centered at a target frequency is generated by coherently summing each of the output signals output by the square law detectors.

Abstract: A method for classifying at least one object of interest in a video is provided. The method includes accessing, using at least one processing device, a frame of the video, the frame including at least one object of interest to be classified, performing, using the at least one processing device, object detection on the frame to detect the object of interest, tracking, using the at least one processing device, the object of interest over a plurality of frames in the video over time using a persistent tracking capability, isolating, using the at least one processing device, a segment of the frame that includes the object of interest, classifying, using the at least one processing device, the object of interest by processing the segment using deep learning, and generating an output that indicates the classification of the object of interest.

Abstract: A method for classifying at least one object of interest in a video is provided. The method includes accessing, using at least one processing device, a frame of the video, the frame including at least one object of interest to be classified, performing, using the at least one processing device, object detection on the frame to detect the object of interest, tracking, using the at least one processing device, the object of interest over a plurality of frames in the video over time using a persistent tracking capability, isolating, using the at least one processing device, a segment of the frame that includes the object of interest, classifying, using the at least one processing device, the object of interest by processing the segment using deep learning, and generating an output that indicates the classification of the object of interest.

Abstract: A method and apparatus for detecting an energetic material on a surface. Electromagnetic radiation is generated. A beam of the electromagnetic radiation is transmitted from a handheld device towards the surface. Reflected electromagnetic radiation from the beam reflecting off of the surface is detected at the handheld device to generate test data. The test data is analyzed at a base unit that is separate from the handheld device to determine whether the energetic material is present on the surface.

Abstract: A method for classifying at least one object of interest in a video is provided. The method includes accessing, using at least one processing device, a frame of the video, the frame including at least one object of interest to be classified, performing, using the at least one processing device, object detection on the frame to detect the object of interest, tracking, using the at least one processing device, the object of interest over a plurality of frames in the video over time using a persistent tracking capability, isolating, using the at least one processing device, a segment of the frame that includes the object of interest, classifying, using the at least one processing device, the object of interest by processing the segment using deep learning, and generating an output that indicates the classification of the object of interest.

Abstract: A method and apparatus for detecting an energetic material on a surface. Electromagnetic radiation is generated. A beam of the electromagnetic radiation is transmitted from a handheld device towards the surface. Reflected electromagnetic radiation from the beam reflecting off of the surface is detected at the handheld device to generate test data. The test data is analyzed at a base unit that is separate from the handheld device to determine whether the energetic material is present on the surface.

Abstract: A coherent fiber array is used to optically detect a radiometric event. The coherent fiber array has a dome-shaped detection surface and a planar output surface. Optical energy from the radiometric event is detected at the dome-shaped detection surface and transferred to the output surface. The coherent fiber array retains directionality of the radiometric event while transferring the optical energy from the dome-shaped detection surface to the planar output surface.

Abstract: A coherent fiber array is used to optically detect a radiometric event. The coherent fiber array has a dome-shaped detection surface and a planar output surface. Optical energy from the radiometric event is detected at the dome-shaped detection surface and transferred to the output surface. The coherent fiber array retains directionality of the radiometric event while transferring the optical energy from the dome-shaped detection surface to the planar output surface.