Abstract: A cloth structure includes a first weaving layer, a second weaving layer interwoven with the first weaving layer, an intermediate connecting layer located between the first weaving layer and the second weaving layer, and a sun protection layer applied on the first weaving layer. The cloth structure presents a single-cloth fabric. The sun protection layer covers pores of the first weaving layer and forms a first protection layer to filter the smog. The first weaving layer is formed with inverted V-shaped cloth patterns, so that the first weaving layer forms a second protection layer to filter the smog. The intermediate connecting layer forms a third protection layer to filter the smog. The second weaving layer obstructs the thread openings of the first weaving layer and forms a fourth protection layer to filter the smog.

Abstract: A method of manufacturing a semiconductor structure includes receiving a first substrate including an IMD layer disposed over the first substrate and a plurality of conductive bumps disposed in the IMD layer; receiving a second substrate; disposing a patterned adhesive over the first substrate, wherein at least a portion of the IMD layer is exposed through the patterned adhesive; and bonding the first substrate with the second substrate, wherein a top surface of the at least portion of the IMD layer is exposed through the patterned adhesive after bonding the first substrate with the second substrate.

Abstract: A machine learning method and a machine learning device are provided. The machine learning method includes: receiving an input signal and performing normalization on the input signal; transmitting the normalized input signal to a convolutional layer; and adding a sparse coding layer after the convolutional layer, wherein the sparse coding layer uses dictionary atoms to reconstruct signals on a projection of the normalized input signal passing through the convolutional layer, and the sparse coding layer receives a mini-batch input to refresh the dictionary atoms.

Abstract: A recycled fancy yarn includes at least two core strands forming a warp thread, and at least one ornamental strand that is twined, twisted or wound around the warp thread to form a weft thread. The core strand is formed by a single first monofilament or a single first filament bundle, and the ornamental strand is formed by a single second monofilament or a single second filament bundle. The first monofilament and/or second monofilament is made of an environmentally friendly material; or the recycled fancy yarn can be recycled or naturally decomposed in the environment, such that the quantity of produced wastes can be reduced, and environmental burden is hardly caused.

Abstract: A negative electrode of a thin film battery and method for forming the same, wherein the negative electrode comprises a porous structural layer, a capacitor layer, and a lithium ion source layer. The porous structural layer is formed on a metal substrate, and a thickness of the porous structural layer is between 200 nm and 700 nm. The capacitor layer is formed on the porous structural layer, and a thickness is between 100 nm and 300 nm. The lithium ion source layer is formed on the capacitor layer. Since the porous structural layer is made of stable material, a problem of charging-discharging instability that is occurred due to damage of battery structure caused by the volume expansion of the capacitor layer during the charging-discharging process can be improved. In addition, the negative electrode can be combined with a positive electrode for forming a thin film battery.

Abstract: An absolute position readout apparatus includes an encoder device and a readout device. The readout device includes multiple first and second magnetic sensing components that correspond to an absolute track of the encoder device, and a third magnetic sensing component and a fourth magnetic sensing components that correspond to an incremental track of the encoder device. The third magnetic sensing component is configured to be spaced apart from the fourth magnetic sensing component by a specific distance, so as to prevent misreading of absolute position information from the first or second magnetic sensing components being at positions corresponding to boundaries between adjacent magnetized regions of the absolute track.

Abstract: Systems and methods for a local proxy for service discovery. In some embodiments, an Internet-of-Things (IoT) gateway may include: a processor; and a memory coupled to the processor, the memory including program instructions stored thereon that, upon execution by the processor, cause the IoT gateway to: retrieve, by a service discovery agent, endpoint information maintained by a service discovery server remotely located with respect to the IoT gateway; store the service endpoint information in the memory; receive a service request aimed at a service endpoint; and provide the service endpoint information, from the memory, in response to the service request.

Abstract: A method for forming a semiconductor device structure is provided. The method includes forming a first overlay grating over a substrate. The first overlay grating has a first strip portion and a second strip portion. The method includes forming a first layer over the first overlay grating. The first layer has a first trench elongated in a second elongated axis, the second elongated axis is substantially perpendicular to the first elongated axis, and the first trench extends across the first strip portion and the second strip portion. The method includes forming a second overlay grating over the first layer. The second overlay grating has a third strip portion and a fourth strip portion. The third strip portion and the fourth strip portion are elongated in the first elongated axis and are spaced apart from each other.

Abstract: Embodiments provide a packaging arrangement that includes a high density interconnect bridge for interconnecting dies within the packaging arrangement. The packaging arrangement comprises one or more redistribution layers and an interconnect bridge embedded within the one or more redistribution layers. A first die is coupled to (i) a first portion of the one or more redistribution layers and (ii) a first portion of the interconnect bridge. A second die coupled to a (ii) a second portion of the one or more redistribution layers and (ii) a second portion of the interconnect bridge to electrically couple the first die and the second die via at least the first interconnect bridge.

Abstract: The present disclosure describes apparatuses and techniques for self-aligning integrated circuit (IC) dies. In some aspects, a hydrophobic material is deposited on a surface of a substrate to form a pattern on the surface of the substrate. The pattern may expose areas of the substrate surface for placement of IC dies. A water-based solution is then applied to the exposed areas such that droplets form on the exposed areas of the substrate surface. IC dies are placed on the droplets of the water-based solution, which can cause the IC dies to align with the exposed areas of the substrate surface. The droplets are then caused to evaporate such that the IC dies settle on the exposed areas of the substrate surface.

Abstract: Systems and methods are disclosed for extending and reinforcing a knowledge graph using wiki-like web pages as a source of information. A web crawler parse a wiki-like source and obtain a topic entity from the source. Relationships between the topic entity and sub-topics within the source are identified and a graph is built with the topic and relationships to sub-topics. A candidate topic in the knowledge graph is identified, and a sub-graph of the knowledge graph is either identified or generated. The knowledge sub-graph contains the candidate topic and relationships to sub-topics. A similarity is computed between the source graph and the knowledge sub-graph. If the two graphs are sufficiently similar, then the source topic graph is merged with the knowledge graph.

Abstract: A decoding method for a rewritable non-volatile memory module is provided. The method includes reading data from a plurality of memory cells of the rewritable non-volatile memory module according to a first voltage, wherein the data includes a user data string and an error checking and correcting code set. The method also includes decoding at least part of sub data units i the user data string according to a first decoding algorithm to obtain a plurality of decoded sub data units. The method further includes restoring a value of the corrected bit to an original bit value if a corrected bit in the decoded sub data units matches a reliability condition.

Abstract: This invention discloses a hair protection effect of vanillyl alkyl ether and the application thereof. The mentioned application comprises the process of topically applying a composition comprising vanillyl alkyl ether compound to reduce hair loss.

Abstract: A decoding method, a memory storage device and a memory control circuit unit are provided. The method includes: arranging a plurality of first voltage levels which are used continuously for reading first data from a plurality of first memory cells according to a first wear degree of the first memory cells; decoding the first data which is read by the arranged first voltage levels; arranging a plurality of second voltage levels which are used continuously for reading second data from the first memory cells according to a second wear degree of the first memory cells, wherein the first wear degree of the first memory cells is different from the second wear degree of the first memory cells, and a voltage gap between any two neighboring voltages levels among the first voltage levels is different from a voltage gap between any two neighboring voltage levels among the second voltage levels; and decoding the second data which is read by the arranged second voltage levels.

Abstract: A source separation method and a source separation device are provided. The source separation method comprises: obtaining at least two source time-frequency signals and a mixed time-frequency signal of the at least two source time-frequency signals; disposing the mixed time-frequency signal at an input layer of a complex-valued deep neural network, and taking the at least two time-frequency signals as a target of the complex-valued deep neural network; calculating a cost function of the complex-valued deep neural network; and performing partial differential to a real part and an imaginary part of a network parameter of the complex-valued deep neural network respectively to minimize the cost function.

Abstract: Systems and methods are disclosed to identify entities that have a similar meaning, and may, in embodiments, be grouped into entity groups for knowledge base construction. In embodiments, the entity relations of similarity or non-similarity for an entity pair are predicted as a binary relationship. In embodiments, the prediction may be based upon similarity score between the entities and the entity features, which features are constructed using an entity feature or representation model. In embodiments, the prediction may be an iterative process involving minimum human checking and existing knowledge update. In embodiments, one or more entity groups are formed using graph search from the predicted entity pairs. In embodiments, a group centroid entity may be selected to represent each group based on one or more factors, such as its generality or popularity.

Abstract: The present disclosure provides a printed circuit board connecting structure including a first printed circuit board, a protection layer, an ACF (anisotropy conductive film), and a second printed circuit board. The first printed circuit board includes a first circuit. The protection layer covers a first area of the first circuit. The ACF covers a second area of the first circuit and extending to the protection layer. The second printed circuit board includes a second circuit facing the first circuit. The second printed circuit board partially overlaps with the ACF on the second area of the first circuit, and there is a gap between the second printed circuit board and the protection layer.

Abstract: A display device includes a display module and an optical film. The display module has a through hole. At least one optical film is located on the display module and has an alignment through hole aligned with the through hole and an alignment mark surrounding the alignment through hole.

Abstract: A semiconductor wafer is singulated to form a plurality of semiconductor packages. The semiconductor wafer has a semiconductor substrate, a metal layer, an adhesive layer, a rigid supporting layer, a passivation layer and a plurality of contact pads. A semiconductor package has a semiconductor substrate, a metal layer, an adhesive layer, a rigid supporting layer, a passivation layer and a plurality of contact pads. A thickness of the rigid supporting layer is larger than a thickness of the semiconductor substrate. A thickness of the metal layer is thinner than the thickness of the semiconductor substrate. An entirety of the rigid supporting layer may be made of a single crystal silicon material or a poly-crystal silicon material. The single crystal silicon material or the poly-crystal silicon material may be fabricated from a reclaimed silicon wafer. An advantage of using a reclaimed silicon wafer is for a cost reduction.

Abstract: The present invention discloses a method of treating cancer cells. The method comprises administrating a composition comprises a rhodamine or rhodamine derivative conjugated to a peptide comprises an amino acid sequence of SEQ ID NO: 1 to a subject in need, and activate the composition with light.