Abstract: The present disclosure provides systems and methods that combine physics-based systems with machine learning to generate synthetic LiDAR data that accurately mimics a real-world LiDAR sensor system. In particular, aspects of the present disclosure combine physics-based rendering with machine-learned models such as deep neural networks to simulate both the geometry and intensity of the LiDAR sensor. As one example, a physics-based ray casting approach can be used on a three-dimensional map of an environment to generate an initial three-dimensional point cloud that mimics LiDAR data. According to an aspect of the present disclosure, a machine-learned model can predict one or more dropout probabilities for one or more of the points in the initial three-dimensional point cloud, thereby generating an adjusted three-dimensional point cloud which more realistically simulates real-world LiDAR data.

Abstract: Disclosed herein is a chip package and method for fabricating the same are provided that includes a redistribution layer (RDL) with a plurality of loop and void structures. The chip package includes an integrated circuit (IC) die, and a package substrate. The RDL is disposed between the IC die and the package substrate. The RDL has RDL circuitry that connects the IC die to the package substrate. The RDL circuitry includes a first coil formed in a first metal layer and a second coil formed in a second metal layer. A first end of the second coil is coupled to a second end of the first coil by a first via. A second end of the second coil is the IC die.

Abstract: A workpiece holder includes a chuck body and a seal ring. The chuck body includes a receiving surface configured to receive a workpiece and at least one vacuum port configured to apply a vacuum seal. The seal ring surrounds a side surface of the chuck body. A top surface of the seal ring is higher than the receiving surface of the chuck body, and the workpiece leans against the seal ring when the vacuum seal is applied between the workpiece and the chuck body.

Abstract: A micro light-emitting diode display panel including first and second substrates, micro light-emitting diodes, a wavelength conversion layer, a light-shielding pattern layer, a light filter layer, and an air gap is provided. The micro light-emitting diodes are disposed on the first substrate and respectively located in a plurality of sub-pixel areas. The micro light-emitting diodes are adapted to emit a light beam. The wavelength conversion layer is overlapped with at least a portion of the micro light-emitting diodes. The light beam is used to excite the wavelength conversion layer to emit a converted light beam. The light filter layer is disposed between the wavelength conversion layer and the second substrate and overlapped with the micro light-emitting diodes. The air gap is disposed between any two adjacent ones of any one of the micro light-emitting diodes, the second substrate, the wavelength conversion layer, and the light filter layer.

Abstract: A method of forming a semiconductor device includes forming a source/drain region on a substrate and forming a first interlayer dielectric (ILD) layer over the source/drain region. The method further includes forming a second ILD layer over the first ILD layer, forming a source/drain contact structure within the first ILD layer and the second ILD layer, and selectively removing a portion of the source/drain contact structure to form a concave top surface of the source/drain contact structure.

Abstract: A joystick module includes a casing, a movable component, a circuit board, a base, a swing arm, a joystick and a sensor. The movable component is disposed inside or outside the casing. The base is disposed within the casing. The swing arm is disposed within the casing, pivotally connected to the base and connected to the movable component for driving the movable component to move. The joystick is connected to the swing arm for driving the swing arm to move. The sensor is disposed on the circuit board and opposite to the movable component, and configured to sense a plurality of received signals from the movable component. The received signals are different with the movement of the movable component.

Abstract: Generally, the disclosed systems and methods implement improved detection of objects in three-dimensional (3D) space. More particularly, an improved 3D object detection system can exploit continuous fusion of multiple sensors and/or integrated geographic prior map data to enhance effectiveness and robustness of object detection in applications such as autonomous driving. In some implementations, geographic prior data (e.g., geometric ground and/or semantic road features) can be exploited to enhance three-dimensional object detection for autonomous vehicle applications. In some implementations, object detection systems and methods can be improved based on dynamic utilization of multiple sensor modalities. More particularly, an improved 3D object detection system can exploit both LIDAR systems and cameras to perform very accurate localization of objects within three-dimensional space relative to an autonomous vehicle.

Abstract: A garden sprinkler has an input base, an output member rotatably installed at the top of the input base, and a rotating member capable of driving the output member. The output Both sides of the output member are respectively equipped with a first outlet and a second outlet. The first outlet is equipped with a first sprinkler, and the second outlet is equipped with a second sprinkler and an adjusting sleeve. The second sprinkler is a side outlet and is screwed with the adjusting sleeve. The second sprinkler can be adjusted by the adjusting sleeve, which improves the convenience of sprinkling operation without multiple sprinkler heads.

Abstract: A package structure includes a first package, a second package, a conductive spacer, and a flux portion. The first package includes a semiconductor die. The second package is stacked to the first package. The conductive spacer is disposed between and electrically couples the first package and the second package. The flux portion is disposed between and electrically couples the first package and the conductive spacer, where the flux portion includes a first portion and a second portion separating from the first portion by a gap, and the first portion and the second portion are symmetric about an extending direction of the gap. The gap is overlapped with the conductive spacer.

Abstract: A controlling system of an endoscopic surgical robot based on head tracking is disclosed. The system includes a host, the robot and a complex camera. The robot holds an endoscope at a specific position through a specific gesture. The complex camera includes a 2D camera for obtaining facial features of a doctor to identify whether the doctor is a registered user. The complex camera includes a 3D camera for obtaining depth information of the doctor's head. The host estimates a 2D space information and a 3D space information of the doctor's head according to the facial features and the depth information, and transforms the 2D space information and 3D space information to a robot moving information by a transformation matrix to control the robot to move.

Abstract: Systems and methods for optical communication are provided. For instance, a method for optical communication can include receiving, by a first coupling module, a power-on signal from a first electronic device coupled to the first coupling module. The method can also include relaying, by the first coupling module, a first optical signal to a second coupling module coupled to a second electronic device. The method can also include relaying, by the second coupling module, in response to receipt of the first optical signal, a second optical signal to the first coupling module. The method can also include activating, by the first coupling module, in response to receipt of the second optical signal, a data transfer circuit for relaying data via an optical communication interface between the first coupling module and the second coupling module.

Abstract: A backlight module includes a light guide plate and a plurality of light-emitting elements. The light guide plate includes a light-incident surface, an effective light-exiting region, a transition region located between the light-incident surface and the effective light-exiting region, and a light-blocking opening disposed in the transition region. The light-emitting elements are disposed at the light-incident surface and emit light toward the effective light-exiting region. The light-blocking opening is located between a plurality of light-emitting regions formed in the transition region by the plurality of light-emitting elements respectively.

Abstract: The present disclosure describes a method for forming a silicon-based, carbon-rich, low-k ILD layer with a carbon concentration between about 15 atomic % and about 20 atomic %. For example, the method includes depositing a dielectric layer, over a substrate, with a dielectric material having a dielectric constant below 3.9 and a carbon atomic concentration between about 15% and about 20%; exposing the dielectric layer to a thermal process configured to outgas the dielectric material; etching the dielectric layer to form openings; and filling the openings with a conductive material to form conductive structures.

Abstract: Disclosed herein are methods and systems for performing instance segmentation that can provide improved estimation of object boundaries. Implementations can include a machine-learned segmentation model trained to estimate an initial object boundary based on a truncated signed distance function (TSDF) generated by the model. The model can also generate outputs for optimizing the TSDF over a series of iterations to produce a final TSDF that can be used to determine the segmentation mask.

Abstract: The present disclosure provides systems and methods that combine physics-based systems with machine learning to generate synthetic LiDAR data that accurately mimics a real-world LiDAR sensor system. In particular, aspects of the present disclosure combine physics-based rendering with machine-learned models such as deep neural networks to simulate both the geometry and intensity of the LiDAR sensor. As one example, a physics-based ray casting approach can be used on a three-dimensional map of an environment to generate an initial three-dimensional point cloud that mimics LiDAR data. According to an aspect of the present disclosure, a machine-learned geometry model can predict one or more adjusted depths for one or more of the points in the initial three-dimensional point cloud, thereby generating an adjusted three-dimensional point cloud which more realistically simulates real-world LiDAR data.

Abstract: A communication device for detecting an object includes a power module, an antenna array, a first sensor pad, a second sensor pad, and a control unit. The antenna array is excited by the power module, and is configured to provide a first beam group and a second beam group. The first sensor pad is disposed adjacent to the first side of the antenna array. A first capacitance is formed between the first sensor pad and the object. The second sensor pad is disposed adjacent to the second side of the antenna array. A second capacitance is formed between the second sensor pad and the object. The control unit controls the power module according to the first capacitance and the second capacitance, so as to selectively apply at least one power backoff operation to the first beam group and/or the second beam group.

Abstract: A magnetic memory element has a Ru hard mask layer. The use of Ru advantageously allows for closer spacing of adjacent magnetic memory elements leading to increased data density. In addition, the use of Ru as a hard mask reduces parasitic electrical resistance by virtue of the fact that Ru does not oxidize in ordinary manufacturing environments. The magnetic memory element can be formed by depositing a plurality of memory element layers, depositing a Ru hard mask layer, depositing a RIEable layer over the Ru hard mask layer, and forming a photoresist mask over the hard mask layer. A reactive ion etching can be performed to transfer the image of the photoresist mask onto the RIEable layer to form a RIEable mask. An ion etching can then be performed to transfer the image of the RIAable mask onto the underlying Ru hard mask and underlying memory element layers.

Abstract: A method comprises embedding a semiconductor structure in a molding compound layer, depositing a plurality of photo-sensitive material layers over the molding compound layer, developing the plurality of photo-sensitive material layers to form a plurality of openings, wherein a first portion and a second portion of an opening of the plurality of openings are formed in different photo-sensitive material layers and filling the first portion and the second portion of the opening with a conductive material to form a first via in the first portion and a first redistribution layer in the second portion.

Abstract: A method of modifying a layout for an integrated circuit (IC) includes: selecting, in the layout, a circuit region to be scaled; setting a fixed area including a fixed feature in the selected circuit region; and scaling the selected circuit region, without scaling the fixed area including the fixed feature, to obtain a modified layout for the IC.

Abstract: The present disclosure relates to a transgenic plant cell comprising polynucleotide sequences encoding glycolate dehydrogenase, malate synthase, and an inhibitory polynucleotide targeting an endogenous glycolate transporter Plgg1, wherein expression of endogenous glycolate transporter Plgg1 in the transgenic plant cell is about 20% to 80% of expression of endogenous glycolate transporter Plgg1 in a plant cell that is not transformed with an inhibitory polynucleotide targeting an endogenous glycolate transporter Plgg1. Also disclosed are transgenic plants, transgenic plant cultures, and methods for increasing photosynthesis efficiency in plants. The disclosed methods enhance biomass productivity and reduce the negative impact of photorespiration and introduction of transgenic constructs on plant growth.