Abstract: A method and system for rapid retrieval of target images based on artificial intelligence, obtaining a template image and a plurality of known labels corresponding to the template image; extracting an image to be detected from a target image database; inputting both the image to be detected and the template image into a trained convolutional neural network, and outputting a hash code of the image to be detected and a hash code of the template image; obtaining a similarity between the images based on a Hamming distance between the hash codes, then selecting one or more images to be detected with the similarity higher than a set threshold as a retrieval result to output. Accordingly, the method and system is able to better cope with the retrieval of items in complex scenarios.

Abstract: The present disclosure relates to improving recommendations for small shops on an ecommerce platform while maintaining accuracy for larger shops. The improvement is achieved by retraining a machine-learning recommendation model to reduce sample selection bias using a meta-learning process. The retraining process comprises identifying a sample subset of shops on the ecommerce platform, where the sample batch includes shops in both a small size category and shops in a large size category. The machine-learning model is then used to make shop-specific user-item interaction predictions for each shop in the sample batch. A shop-specific loss is calculated for each shop based on the model's predicted user-item interactions for the shop-specific training dataset and the actual user-item interactions in the shop-specific training dataset. A global loss is calculated based on each of the shop-specific losses, and the baseline model is updated to minimize the global loss.

Abstract: A speaker includes: a frame, a vibration system, including a diaphragm, a voice coil, and a flexible circuit board, the flexible circuit board including a first fixing portion, a second fixing portion, and an elastic arm, and a magnetic circuit system, including a yoke with a through hole and a main magnet and a secondary magnet fixed to the yoke, the secondary magnet surrounding and spaced from the main magnet for forming a magnetic gap, the voice coil located in the magnetic gap, wherein a projection of the elastic arm along a vibration direction of the diaphragm is at least partially located in the through hole, the speaker further comprises a connecting piece connected with an outer surface of the yoke, the connecting piece covers the through hole. Compared with the related art, the speaker disclosed by the present disclosure has better acoustic performance.

Abstract: A multiple scenario-oriented item retrieval method and system. The method includes the steps of extracting, by Hashing learning, image features from an image training set to train a pre-built item retrieval model; when an image is in a scenario of hard samples, introducing an adaptive similarity matrix, optimizing the similarity matrix by an image transfer matrix, constructing an adaptive similarity matrix objective function in combination with an image category label; constructing a loss quantization objective function between the image and a Hash code according to the image transfer matrix; when the image is in a scenario of zero samples, introducing an asymmetric similarity matrix, constructing an objective function by taking the image category label as supervisory information in combination with equilibrium and decorrelation constraints of the Hash code; and training the item retrieval model based on the above objective function to obtain a retrieved result of a target item image.

Abstract: The present disclosure provides a display device comprising: a display panel having a light-exiting upper surface, a lower surface opposite to the light-exiting upper surface and a plurality of side edge end surfaces between the light-exiting upper surface and the lower surface, and having a first group of bonding electrodes at a side where a first side edge end surface in the plurality of side edge end surfaces is located; a first flexible circuit board electrically connected to the first group of bonding electrodes; a first insulating protection layer at a side of the first flexible circuit board away from the first group of bonding electrodes, wherein its upper edge is higher than an upper edge of the first flexible circuit board, and its lower edge is lower than the first group of bonding electrodes; a first conductive layer; and a first adhesive layer. The present disclosure also provides a display apparatus comprising the display device and a composite adhesive tape for a display device.

Abstract: A doped cathode material for lithium-ion batteries is disclosed. Methods and systems are further provided for doping a cathode material for use in a lithium-ion battery. In one example, the doping may be a dry surface doping process. In some examples, dopants may stabilize a crystal structure of the cathode material and may result in fewer side reactions with an electrolyte as compared to an undoped cathode material. As such, cycling performance and capacity retention may be improved relative to the undoped cathode material. Further, in some examples, the doped cathode material produced with the dry surface doping process may have improved cycling performance and capacity retention relative to a comparable doped cathode material produced with a wet surface doping process.

Abstract: The present disclosure provides a speaker unit, which includes a frame, and a vibration system and a magnetic circuit system fixed on the frame. The vibration system includes a diaphragm and a voice coil that drives the diaphragm to vibrate and sound. The speaker unit further includes a magnetic front cover fixed on the frame for improving the BL value of the voice coil. The speaker unit of the present disclosure improves the BL value of the voice coil.

Abstract: A leveling foot assembly includes a foot and a rod secured to the foot. The rod includes a first threaded portion, a second threaded portion, and a third threadless portion positioned between the first threaded portion and the second threaded portion. The leveling foot assembly further includes a leveling nut configured to engage the first threaded portion of the rod when the foot is in a raised position and configured to engage the second threaded portion of the nut when the foot is in a lowered position.

Abstract: The present disclosure relates to improving recommendations for small shops on an ecommerce platform while maintaining accuracy for larger shops. The improvement is achieved by retraining a machine-learning recommendation model to reduce sample selection bias using a meta-learning process. The retraining process comprises identifying a sample subset of shops on the ecommerce platform, and then creating shop-specific versions of the recommendation model for each of the shops in the subset. Each shop-specific model is created by optimizing the baseline model to predict user-item interactions in a first training dataset for the applicable shop. Each of the shop-specific models is then tested using a second training dataset for the shop. A loss is calculated for each shop-specific model based on the model's predicted user-item interactions and the actual user-item interactions in the second training dataset for the shop.

Abstract: A speaker includes: a frame, a vibration system, including a diaphragm and a voice coil driving the diaphragm to vibrate and emit sounds, and a magnetic circuit system fixed to the frame, including a lower plate, and a central magnetic portion and an auxiliary magnet fixed to the lower plate, the central magnetic portion spaced apart from the auxiliary magnet, the auxiliary magnet surrounding the central magnetic portion and forming a magnetic gap, the voice coil located in the magnetic gap, wherein the central magnetic portion includes a main magnetic portion and a reinforcing portion surrounding the main magnetic portion, a projection of the voice coil in a vibration direction of the diaphragm is at least partially located in the reinforcing portion. Compared with the related art, the speaker disclosed by the present disclosure could improve the BL value of the voice coil.

Abstract: The present disclosure relates to a multi-stage temporary plugging fracturing method based on a stress cage effect, a device thereof and a storage medium, which are applied to the technical field of oil and gas exploitation. The method includes: determining a plugging point by acquiring a weak stress position on a major fracture; determining the consumption of temporary plugging deflection fluid by calculating a net pressure in a fracture that needs to be increased to generate a branch deflection fracture at the position where a principal stress is minimum; plugging a weakest stress position in the major fracture to form an artificial shielding layer, preventing subsequent fluid pressure from passing forward, thereby changing a geostress field near this position, thus forming an additional induced stress field near the weak position, generating a stress cage effect, and realizing formation of a multi-stage temporary plugging branch fracture in vertical wells.

Abstract: A doped cathode material for lithium-ion batteries is disclosed. Methods and systems are further provided for doping a cathode material for use in a lithium-ion battery. In one example, the doping may be a dry surface doping process. In some examples, dopants may stabilize a crystal structure of the cathode material and may result in fewer side reactions with an electrolyte as compared to an undoped cathode material. As such, cycling performance and capacity retention may be improved relative to the undoped cathode material. Further, in some examples, the doped cathode material produced with the dry surface doping process may have improved cycling performance and capacity retention relative to a comparable doped cathode material produced with a wet surface doping process.

Abstract: The present disclosure discloses a sound device includes a frame having a receiving space, a vibration system and a magnet system having a magnetic gap; the vibration system includes a diaphragm and a coil; the magnet system includes a yoke and a magnet; the sound device further includes a FPCB connecting the coil with an outer circuit; the FPCB includes a first welding portion connecting with the coil, a second welding portion connecting with the outer circuit, and a connection portion connecting the first welding portion and the second welding portion, the second welding portion is fixed to the frame and located outside the receiving space. The sound device in the present disclosure has higher space utilization rate.

Abstract: A method for direct recycling of degraded lithium-ion battery (LIB) cathodes includes relithiating degraded lithium nickel cobalt manganese oxide (NCM) by mixing the cathode particles with a eutectic molten-salt solution and heating the mixture at ambient pressure over a period of time, followed by a short-time thermal annealing. Combining low-temperature relithiation using the eutectic molten-salt solution with thermal annealing provides successful regeneration and full recovery of the LIB to its original storage capacity, cycling stability and rate capability to the levels of the pristine materials. The method is useful to recycle and remanufacture degraded cathode materials for LIB or sodium-ion batteries.

Abstract: A method of preparing lithium nickel manganese cobalt oxide high-nickel single-crystal positive electrode material comprises the following steps: mixing and ball milling nickel cobalt manganese hydroxide with a lithium source, zirconium oxide, tungsten oxide, and sodium carbonate, then performing primary sintering, pulverizing, and then obtaining an lithium nickel manganese cobalt oxide high-nickel single-crystal positive electrode material intermediate, the temperature of the primary sintering being 50-150° C. higher than a normal sintering temperature; and evenly mixing the lithium nickel manganese cobalt oxide high-nickel single-crystal positive electrode material intermediate and a coating agent, and then performing secondary sintering to obtain the lithium nickel manganese cobalt oxide high-nickel single-crystal positive electrode material.

Abstract: A speaker includes: a frame, a vibration system, including a diaphragm and a voice coil assembly, and a magnetic circuit system, comprising a lower plate, a first main magnet and a first auxiliary magnet, a main pole plate, and an upper plate, wherein the upper plate includes a body portion and a groove recessed from the body portion to the lower plate, the groove includes a base, and a connecting arm connected with the body portion and the base, the base is closer to the body portion than the lower plate, the speaker further includes a second auxiliary magnet located in the groove, the base is clamped between the first auxiliary magnet and the second auxiliary magnet.

Abstract: The present disclosure relates to a method and a device for predicting ability of a temporary plugging agent to plug cracks and a storage medium, and includes the following steps: calculating a plane elastic modulus of a core through experiments, acquiring a maximum force load of a specimen, calculating rock fracture toughness through the maximum force load, manufacturing an artificial crack model and placing it in a core holder, preparing temporary plugging deflection fluid by selecting a temporary plugging agent, injecting the temporary plugging deflection fluid into the artificial crack so as to calculate apparent fracture toughness of the temporarily plugged crack, and finally calculating a fracture pressure of the temporarily plugged crack based on the calculated rock fracture toughness and the plane elastic modulus of the core. Through the above method, the present disclosure provides a standard fracture pressure calculation method.

Abstract: A large-flux in-situ filtration device for marine microorganism based on a multi-channel circulation distributor is provided, comprising an energy and control module, a circulation distribution module, a filtration module, a pump module, a filtration volume measurement module, a pipeline connection, and a main body support frame. The circulation distribution module may include a multi-channel circulation distributor, a deep-sea servo motor, etc. The filtration module may include a plurality of filter film holders. The pump module may include a stainless steel gear pump, a deep-sea drive motor, etc. The filtration volume measurement module may include a water meter, a camera, and an LED light.

Abstract: An ash collecting system includes an ash discharge device, an ash conveying pipeline and an ash collecting tank. A first clean air chamber and a first foul air chamber are arranged in a shell. A plurality of first filtering structures and vibrating structures are arranged in the first clean air chamber. The first foul air chamber is provided with a sweeping structure for sweeping ash. The ash collecting tank includes a cylindrical shell and an ash bin. The ash removal port is in communication with the second air inlet. A second clean air chamber and a second foul air chamber are arranged in the cylindrical shell. Second filtering structures are arranged in the second foul air chamber. The ash bin is in communication with the second foul air chamber. An ash collecting structure for collecting ash is arranged in the ash bin.

Abstract: The present disclosure relates to improving recommendations for small shops on an ecommerce platform while maintaining accuracy for larger shops by retraining a machine-learning recommendation model to reduce sample selection bias using a meta-learning process. The retraining process comprises identifying a sample subset of shops on the ecommerce platform, and then creating shop-specific versions of the recommendation model for each of the shops in the subset by optimizing the baseline model to predict user-item interactions in a first training dataset for the applicable shop. Each of the shop-specific models is then tested using a second training dataset for the shop. A loss is calculated for each shop-specific model based on the model's predicted user-item interactions and the actual user-item interactions in the second training dataset for the shop. A global loss is calculated based on each of the shop-specific losses, and the baseline model is updated to minimize the global loss.