Abstract: A hardware-aware zero-cost neural network architecture search system is configured to perform the following. A neural network search space is divided into multiple search blocks. Each of the search blocks includes multiple candidate blocks. The candidate blocks are guided and scored through a latent pattern generator. The candidate blocks in each of the search blocks are scored through a zero-cost accuracy proxy. One of the candidate blocks included in each of the search blocks is sequentially selected as selected candidate blocks, the selected candidate blocks are combined into multiple neural networks to be evaluated, and network potential of the neural networks to be evaluated is calculated according to scores of the selected candidate blocks. One neural network to be evaluated with the highest network potential is selected to determine the corresponding selected candidate blocks.

Abstract: A chuck vacuum line of a semiconductor processing tool includes a first portion that penetrates a sidewall of a main pumping line of the semiconductor processing tool. The chuck vacuum line includes a second portion that is substantially parallel to the sidewall of the main pumping line and to a direction of flow in the main pumping line. A size of the second portion increases between an inlet end of the second portion and an outlet end of the second portion along the direction of flow in the main pumping line.

Abstract: Aspects of the disclosure provide a method for merging compressed access units according to compression rates and/or positions of the respective compressed access units. The method can include receiving a sequence of compressed access units corresponding to a sequence of raw access units partitioned from an image or a video frame and corresponding to a sequence of memory spaces in a frame buffer, determining a merged access unit including at least two consecutive compressed access units based on compression rates and/or positions of the sequence of compressed access units. The merged access unit is to be stored in the frame buffer with a reduced gap between the at least two consecutive compressed access units compared with storing the at least two consecutive compressed access units in corresponding memory spaces in the sequence of memory spaces.

Abstract: A system and method for anticipatory contextual notifications are provided. In example embodiments, a context indication corresponding to a user that is a member of a social network service is received. The context indication is indicative of a real-time context of the user. The context indication is analyzed to determine that the context indication is a significant context indication. An entity associated with the significant context indication and that is a member of the social network service is identified. Member data associated with the identified entity and the user is accessed. Notification context is generated based, at least in part, on an analysis of the significant context indication and the member data. Presentation, to the user, of a notification that includes the notification context is caused.

Abstract: An energy saving control structure for water-cooled air conditioning includes: a power module; a first electric circuit, configured on one side of the power module and in selective electric connection therewith, the first electric circuit in electric connection with at least one air circulation device; and a second electric circuit, configured on another side of the power module and in selective electric connection therewith, the second electric circuit in electric connection with at least one cooling device and a compressor configured on one side of the cooling device, and the cooling device including a fan motor of a cooling water tower and at least one second motor configured on one side of the fan motor of the cooling water tower.

Abstract: Aspects of the disclosure provide a method for merging compressed access units according to compression rates and/or positions of the respective compressed access units. The method can include receiving a sequence of compressed access units corresponding to a sequence of raw access units partitioned from an image or a video frame and corresponding to a sequence of memory spaces in a frame buffer, determining a merged access unit including at least two consecutive compressed access units based on compression rates and/or positions of the sequence of compressed access units. The merged access unit is to be stored in the frame buffer with a reduced gap between the at least two consecutive compressed access units compared with storing the at least two consecutive compressed access units in corresponding memory spaces in the sequence of memory spaces.

Abstract: A vehicle safety method includes steps: monitoring a face state of a driver and determining whether the face state satisfies a tired state or not; sensing a physiological state of the driver and determining whether the physiological state satisfies a conscious state or not; sending a warning query when the face state satisfies the fired state and the physiological state satisfies the conscious state; determining whether a response corresponding to the warning query is received or not; and activating a vehicle safety mechanism if the response corresponding to the warning query is not received.

Abstract: A system and method for anticipatory contextual notifications are provided. In example embodiments, a context indication corresponding to a user that is a member of a social network service is received. The context indication is indicative of a real-time context of the user. The context indication is analyzed to determine that the context indication is a significant context indication. An entity associated with the significant context indication and that is a member of the social network service is identified. Member data associated with the identified entity and the user is accessed. Notification context is generated based, at least in part, on an analysis of the significant context indication and the member data. Presentation, to the user, of a notification that includes the notification context is caused.

Abstract: A semiconductor light-emitting device includes at least one light-emitting chip. The light-emitting chip includes plural light-emitting units, which are electrically coupled to each other in series, in parallel or in series-parallel combination; a first-type electrode electrically coupled to an external power source, the first-type electrode being disposed on one of the light-emitting units; a second-type electrode disposed on another of the light-emitting units; and a tapped point for electrically coupling at least one of the light-emitting units to an electronic component.

Abstract: An illuminating device includes at least one light-emitting source. The light-emitting source includes a substrate; at least one light-emitting chip disposed on the substrate; and at least one constant-current component electrically coupled to the light-emitting chip. The light-emitting chip includes multiple light-emitting units that are electrically coupled in series, in parallel, or in series-parallel; a first-type electrode, disposed on at least one of the light-emitting units, for electrically coupling to a central DC power source; a second-type electrode disposed on at least one light-emitting unit different from the one, on which the first-type electrode is disposed; and a tapped point configured for electrically coupling at least one of the light-emitting units to the constant-current component.

Abstract: A semiconductor light-emitting device includes at least one light-emitting chip. The light-emitting chip includes plural light-emitting units, which are electrically coupled to each other in series, in parallel or in series-parallel combination; a first-type electrode electrically coupled to an external power source, the first-type electrode being disposed on one of the light-emitting units; a second-type electrode disposed on another of the light-emitting units; and a tapped point for electrically coupling at least one of the light-emitting units to an electronic component.

Abstract: A method and a device for automatic color correction by defining an operation window, a target window, and an active region bounded by the preceding windows so as to adjust chroma values in the active region. Correction of chroma signals is associated with a window region rather than a certain point. Besides, the present invention provides correction of the chroma signals so as to prevent image discontinuity and disharmony.

Abstract: A method and a device for automatic color correction by defining an operation window, a target window, and an active region bounded by the preceding windows so as to adjust chroma values in the active region. Correction of chroma signals is associated with a window region rather than a certain point. Besides, the present invention provides correction of the chroma signals so as to prevent image discontinuity and disharmony.

Abstract: The present infra-red light-emitting device includes a substrate with a first window layer, a silicon dioxide layer positioned on the first window layer, silicon nanocrystals distributed in the silicon dioxide layer, a second window layer, a transparent conductive layer and a first ohmic contact electrode positioned in sequence on the silicon dioxide layer, and a second ohmic contact electrode positioned on the bottom surface of the substrate. The present method forms a sub-stoichiometric silica (SiOx) layer on a substrate, wherein the numerical ratio (x) of oxygen atoms to silicon atoms is smaller than 2. A thermal treating process is then performed in a nitrogen or argon atmosphere to transform the SiOx layer into a silicon dioxide layer with a plurality of silicon nanocrystals distributed therein. The thickness of the silicon dioxide layer is between 1 and 10,000 nanometers, and the diameter of the silicon nanocrystal is between 4 and 8 nanometers.

Abstract: The present red light-emitting device includes a substrate with a first window layer, a silicon dioxide layer positioned on the first window layer, a plurality of silicon nanocrystals distributed in the silicon dioxide layer, a second window layer, a transparent conductive layer and a first ohmic contact electrode positioned in sequence on the silicon dioxide layer, and a second ohmic contact electrode positioned on the bottom surface of the substrate. The present method forms a sub-stoichiometric silica (SiOx) layer on a substrate, wherein the numerical ratio (x) of oxygen atoms to silicon atoms is smaller than 2. A thermal treating process is then performed in an oxygen atmosphere to transform the SiOx layer into a silicon dioxide layer with a plurality of silicon nanocrystals distributed therein. The thickness of the silicon dioxide layer is between 1 and 10,000 nanometers, and the diameter of the silicon nanocrystal is between 3 and 5 nanometers.

Abstract: The present white light-emitting device includes a substrate with first window layer, a trisilicon tetranitride layer positioned on the first window layer, a plurality of silicon nanocrystals distributed in the trisilicon tetranitride layer, a second window layer, a transparent conductive layer and a first ohmic contact electrode positioned in sequence on the trisilicon tetranitride layer, and a second ohmic contact electrode positioned on the bottom surface of the substrate. The present method first forms a sub-stoichiometric silicon nitride (SiNx) layer on a substrate, wherein the numerical ratio (x) of nitrogen atoms to silicon atoms is smaller than 4/3. A thermal treating process is then performed to transform the sub-stoichiometric SiNx layer into a trisilicon tetranitride layer with a plurality of silicon nanocrystals distributed therein. The thickness of the trisilicon tetranitride layer is between 1 and 10,000 nanometers, and the diameter of the silicon nanocrystal is between 1 and 10 nanometers.

Abstract: This present invention describes a performance control method for a computer. An advanced configuration and power interface (ACPI) is used to control the performance of desktop CPU. According to this method, the control items in the ACPI are set according to the throttling function setting and according to the requirement of the user to set the number of the related performance states supported by the CPU. Then, the operation system selects the related performance state of the CPU according to the required system power.

Abstract: This present invention describes a performance control method for a computer. An advanced configuration and power interface (ACPI) is used to control the performance of desktop CPU. According to this method, the control items in the ACPI are set according to the throttling function setting and according to the requirement of the user to set the number of the related performance states supported by the CPU. Then, the operation system selects the related performance state of the CPU according to the required system power.

Abstract: A chinese pictograph or graphic device mounted on a printer port, to which modular design is applied with no need of changing the printer port interface of personal computers PC and printers, this device can be externally mounted, as required, between the printer port interface of a PC and the parallel interface of its printer. With this newly-designed device, the PC can sent forth control signals and status signals to generate pictographic data which can be transmitted to the invented device by way of data bus, and the data, after being processed in the invented device, can be transferred to required pictures or graphics and sent back to the printer port of a PC, thus the PC can read out the required pictures or graphic from its printer port, the PC also can order the printer to print the data which transmitted by data bus no need of the procession of the invented device.