Abstract: A semiconductor device, an integrated fan-out package and a method of forming the same are disclosed. In some embodiments, a semiconductor device includes a substrate, a conductive layer, a passivation layer and a bump structure. The substrate has at least one electronic component therein. The conductive layer has a plurality of lines patterns over and electrically connected to the at least one electronic component. The passivation layer is over the conductive layer. The bump structure has a plurality of protruding parts penetrating through the passivation layer and electrically connected to the lines patterns of the conductive layer.

Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A feature extraction process is performed to obtain features based on the LDR images and the HDR images. A second neural network for video stabilization is trained according to the LDR images and the HDR images based on a loss function by minimizing a loss value of the loss function to generate stabilized HDR images in a time-dependent manner, where the loss value of the loss function depends upon the features. An HDR classifier is constructed according to the LDR images and the HDR images. The stabilized HDR images are classified by using the HDR classifier to generate a reward value, where the loss value of the loss function further depends upon the reward value.

Abstract: An optical device includes a substrate, a single-layer metasurface disposed on the substrate, and a refractive lens. The metasurface and the refractive lens may be configured to bring at least five distinct wavelengths to focus on a same plane.

Abstract: A system for playing specific streaming selected from combined thstreamings and a method thereof are disclosed. In the system, an array server combines video streamings outputted from signal sources, to form a multi-source streaming including all video streamings, and transmits the combined multi-source streaming to a streaming server, the streaming server provides the multi-source streaming to a client end, the client end obtains the video streaming to be played from the received multi-source streaming. Therefore, a user can immediately switch the video streamings with different view angles while watching live broadcast, so as to achieve the technical effect of reducing the streaming load of the server during live broadcast.

Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A second neural network for video stabilization is trained to generate stabilized HDR images in a time-dependent manner.

Abstract: A package structure has a chip, a molding compound encapsulating the chip and an inductor structure disposed above the chip. A vertical projection of the inductor structure at least partially overlaps with a vertical projection of the chip.

Abstract: A display device is provided. The display device includes a panel, a memory, and a controller. The panel includes multiple pixels. The memory includes a first section and a second section. The memory stores an aging record table. Multiple brightness attenuation values recorded in the aging table are respectively divided into multiple first portion attenuation values and multiple second portion attenuation values. The first section stores the first portion attenuation values. The second section stores the second portion attenuation values. The controller includes an update circuit and a compensation circuit. The controller is coupled to the panel and the memory. The update circuit receives gray values displayed by the pixels to update the aging table. The compensation circuit reads the first portion attenuation values from the first section so as to perform an aging compensation on the pixels.

Abstract: A display device is provided. The display device includes a panel, a memory, and a controller. The panel includes multiple pixels. The memory includes a first section and a second section. The memory stores an aging record table. Multiple brightness attenuation values recorded in the aging table are respectively divided into multiple first portion attenuation values and multiple second portion attenuation values. The first section stores the first portion attenuation values. The second section stores the second portion attenuation values. The controller includes an update circuit and a compensation circuit. The controller is coupled to the panel and the memory. The update circuit receives gray values displayed by the pixels to update the aging table. The compensation circuit reads the first portion attenuation values from the first section so as to perform an aging compensation on the pixels.

Abstract: The optical imaging apparatus includes a metasurface lens including a substrate and a plurality of nano-structures patterned on a first side of the substrate. The optical imaging apparatus further includes imaging optics disposed in a spaced apart relationship with a second side of the substrate. The second side is opposite the first side on which the nano-structures are patterned. A surface of the imaging optics and the second side of the substrate define a space for accommodating an immersion fluid. The metasurface lens is configured to direct light incident on the plurality of nano-structures towards the imaging optics through the space accommodating the immersion fluid.

Abstract: A test circuit testing a storage circuit and including a controller, a pattern-generator circuit, a comparing circuit, and a first register is provided. The storage circuit includes a storage block. The controller is configured to generate a plurality of internal test signals. The pattern-generator circuit generates and provides test data to the storage circuit according to the internal test signal. The storage circuit writes the test data into the storage block and reads the storage block to generate read data. The comparing circuit compares the test data and the read data to generate a test result. The first register stores the test result. The controller determines whether the storage circuit is working normally according to the test result stored in the first register.

Abstract: Embodiments of the present disclosure provide an inverter vent and a loudspeaker having the same. Preferably, air flows within the inverted vent in a 360 degrees full-circumferential direction. This design can improve the efficiency of the loudspeaker, reduce the wind noise, and increase the bass ductility.

Abstract: Polarization-insensitive metasurfaces using anisotropic nanostructures are disclosed. These anisotropic structures allow for an accurate implementation of phase, group delay, and group delay dispersion, while simultaneously making it possible to realize a polarizationinsensitive, diffraction-limited and achromatic metalens for wavelength, e.g., ?=from about 460 nm to about 700 nm. The approach of polarization-insensitivity can be also applied for other metasurface devices with applications in, e.g., imaging and virtual or augmented reality.

Abstract: Embodiments of the disclosure provide a package structure and method of forming the same. The package structure includes a first die, a first encapsulant, a first RDL structure, a die stack structure and a second encapsulant. The first encapsulant laterally encapsulates the first die. The first RDL structure is electrically connected to the first die, and disposed on a first side of the first die and the first encapsulant. The die stack structure is electrically connected to the first die and disposed on a second side of the first die opposite to the first side. The second encapsulant is located over the first encapsulant and laterally encapsulating the die stack structure. A sidewall of the first encapsulant is aligned with a sidewall of the second encapsulant.

Abstract: An electric bike having an improved effect of heat dissipation has a body, a driving device, and a battery. The body has a frame, a rear wheel, and a front wheel, wherein the rear wheel and the front wheel are mounted to the frame. The driving device is mounted to the body and has a motor and a shell. The motor is connected with the frame and has a front end near the front wheel and a rear end near the rear wheel. The shell covers the motor and has multiple inlets and multiple outlets disposed therethrough. The multiple inlets are located near the front end of the motor. The multiple outlets are located near the rear end of the motor. The battery is electrically connected to the motor. The front wheel, the battery, the motor, and the rear wheel are serially aligned.

Abstract: A semiconductor chip including a die substrate, a plurality of first bonding structures, a plurality of conductive elements, at least one integrated device, a plurality of conductive posts and a protection layer is provided. The first bonding structures are disposed on the die substrate. The conductive elements are disposed on the die substrate adjacent to the first bonding structures. The integrated device is disposed on the die substrate over the first bonding structures, wherein the integrated device includes a plurality of second bonding structures and a plurality of conductive pillars, and the second bonding structures are hybrid bonded to the first bonding structures. The conductive posts are disposed on the conductive elements and surrounding the integrated device. The protection layer is encapsulating the integrated device and the conductive posts.

Abstract: An ophthalmic lens includes a hybrid plano-convex refractive lens body having a convex portion and a planar portion. A metasurface array can be associated with the planar portion and include an arrangement of metasurface building elements dimensioned from an optical wavelength. The metasurface building elements can be configured across the lens body to define an optical characteristic of the ophthalmic lens. The arrangement of metasurface building elements can include meta-atoms that are configured to induce a polarization-dependent focusing of light received by the ophthalmic lens. A shape of the meta-atoms of the array can be determined based on a function of the ophthalmic lens, including glare/halo reduction. The meta-atoms can be formed as canonical and/or freeform shapes.

Abstract: Packaged semiconductor devices and methods of packaging semiconductor devices are disclosed. In some embodiments, a packaged semiconductor device includes an integrated circuit die, a first molding material disposed around the integrated circuit die, and a through-via disposed within the first molding material. A first side of a redistribution layer (RDL) is coupled to the integrated circuit die, the through-via, and the first molding material. A second molding material is over a second side of the RDL, the second side of the RDL being opposite the first side of the RDL. The packaged semiconductor device includes an antenna over the second molding material.

Abstract: A method and an electronic apparatus of modifying a three-dimensional model are provided. In the method, multiple two-dimensional images are obtained by scanning a real environment Real objects of the real environment are identified in the two-dimensional images. A three-dimensional model is generated based on the two-dimensional images. The three-dimensional model is related to the depth information of the real objects. A modifying operation for modifying one or more virtual objects in the three-dimensional model is received. The virtual objects are corresponding to one or more real objects. The modifying operation is mapped into the two-dimensional images based on the identified result of the real objects in the two-dimensional image and the depth information of the real objects, to generate a modified image. A modified three-dimensional model is generated based on the modified image, to provide a proper modified result.

Abstract: An optical device includes: (1) a substrate; and (2) multiple meta-lenses disposed on the substrate, each meta-lens of the meta-lenses including multiple nanofins disposed on a respective region of the substrate, the nanofins together specifying a phase profile of the meta-lens.

Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A second neural network for video stabilization is trained to generate stabilized HDR images in a time-dependent manner.