Abstract: A surface plasmon infrared nano-pulse laser having a multi-resonance competition mechanism, consisting of the four parts of a surface plasmon nano-pin resonance chamber (1), a spacer layer (2), a gain medium (3), and a two-dimensional material layer (4). The surface plasmon nano-pin resonance chamber (1) consists of a metal nano rod (11) and one or more nano sheets (12) grown thereon, the surface plasmon nano-pin resonance chamber (1) and the gain medium (3) being isolated by the isolating layer (2), and the two-dimensional material layer (4) covering a surface of the surface plasmon nano-pulse laser; positive and negative electrodes (5) are located at two ends of the surface plasmon nano-pulse laser, and a layer of a two-dimensional material having a feature of saturatable absorption is introduced to a surface of the nano-pin resonance chamber.

Abstract: A surface plasmon-semiconductor heterojunction resonant optoelectronic device and a preparation method thereof are provided. A surface ligand molecule is modified on a plasmonic nanostructure, a plasmonic crystal face structure is bound to the surface ligand molecule, a semiconductor nanostructure seed crystal is located on the plasmonic crystal face structure, a one-dimensional semiconductor nanostructure is located on the semiconductor nanostructure seed crystal, and all parts are in tight contact. The heterogeneous integration material achieves a lattice match at an interface, greatly reduces a loss caused by defects and rough crystal faces, and can achieve direct coupling of a surface plasmon mode and an optical mode. The heterogeneous integration material has a large application prospect in the fields of a nanolaser, a nano heat source and photoelectric detection and photocatalysis.

Abstract: A method for host-side data classification according to embodiments, includes: writing a data sector directly from a host to a storage device; storing the data sector in the storage device; sending a write completion acknowledgement from the storage device to the host; in the background: classifying the data sector on the host to estimate an expected lifetime of the data sector; assigning a data type index to the data sector based on the expected lifetime of the data sector; and asynchronously sending the data type index for the data sector from the host to the storage device.

Abstract: Disclosed is a membrane surface modification method. The method is applicable to a variety of hydrophobic membranes by doping selected inorganic particles. One act of the method involves the in-situ embedment of the inorganic particles onto the membrane surface by dispersing the particles in a non-solvent bath for polymer precipitation. Further membrane surface modification can be achieved by hydrothermally growing new inorganic phase on the embedded particles. The embedment of particles is for the subsequent phase growth.

Abstract: A solid-state data storage device according to embodiments includes, a storage device controller; solid-state memory; and an inline encryption engine, embedded in the storage device controller, for encrypting data blocks received from a host using a set of encryption keys and writing the encrypted data blocks into the solid-state memory, wherein data blocks having similar lifetimes are encrypted using the same encryption key.

Abstract: The present invention provides a sound generator, including a magnetic circuit system having a magnet and a lower plate for carrying the magnet. The magnet includes a first surface opposite to the lower plate and a second surface opposite to the first surface. The first surface of the magnet protrudes to form a first bolt, the lower plate is provided with a lower plate jack matching with the first bolt, and the first bolt is inserted into the lower plate jack. The sound generator of the present disclosure improves the structural stability of the magnetic circuit system and is convenient for positioning.

Abstract: A surface plasmon-semiconductor heterojunction resonant optoelectronic device and a preparation method thereof are provided. A surface ligand molecule is modified on a plasmonic nanostructure, a plasmonic crystal face structure is bound to the surface ligand molecule, a semiconductor nanostructure seed crystal is located on the plasmonic crystal face structure, a one-dimensional semiconductor nanostructure is located on the semiconductor nanostructure seed crystal, and all parts are in tight contact. The heterogeneous integration material achieves a lattice match at an interface, greatly reduces a loss caused by defects and rough crystal faces, and can achieve direct coupling of a surface plasmon mode and an optical mode. The heterogeneous integration material has a large application prospect in the fields of a nanolaser, a nano heat source and photoelectric detection and photocatalysis.

Abstract: A host-assisted data deduplication system, including: a storage device including a controller, memory, and a write cache; and a host including a data duplication module coupled to the storage device, wherein the controller of the storage device is configured to write a data sector received from the host into the write cache in the storage device and to send a write completion to the host immediately after writing the data sector to the write cache; wherein the data duplication module is configured to detect whether the data sector is identical to another data sector stored in the memory of the storage device and to asynchronously send a duplicate detection result to the controller of the storage device, and wherein the controller of the storage device is configured to perform a data deduplication process for the data sector stored in the write cache based on the duplicate detection result.

Abstract: A model training method and apparatus for image recognition, and a non-transitory storage medium are provided. The model training method includes: obtaining a multi-label image training set including a plurality of training images each annotated with a plurality of sample labels; selecting target training images from the multi-label image training set for training a current model; performing label prediction on each target training image using the current model, to obtain a plurality of predicted labels of the each target training image; obtaining a cross-entropy loss function corresponding to the plurality of sample labels of the each target training image, a positive label loss being greater than a negative label loss and having a weight greater than 1; converging the predicted labels and the sample labels of the each target training image according to the cross-entropy loss function, and updating parameters of the current model, to obtain a trained model.

Abstract: The present disclosure describes a method, apparatus, and storage medium for generating network representation for a neural network. The method includes obtaining a source-side vector sequence corresponding to an input sequence.

Abstract: Antigen binding molecules (ABMs) comprising Fab domains in non-native configurations, ABM conjugates comprising the ABMs and cytotoxic or cytostatic agents, pharmaceutical compositions containing the ABMs and ABM conjugates, methods of using the ABMs, ABM conjugates and pharmaceutical compositions for treating cancer, nucleic acids encoding the ABMs, cells engineered to express the ABMs, and methods of producing ABMs.

Abstract: A method of establishing an image set for image recognition includes obtaining a single-label image set comprising an image annotated with a single label, and a multi-label image set comprising an image annotated with a plurality of labels; converting content of each label into a corresponding word identifier according to a semantic network, to obtain a word identifier set, a converted single-label image set, and a converted multi-label image set; and constructing a hierarchical semantic structure according to the word identifier set and the semantic network. The method also includes performing label supplementation on the image in the converted single-label image set to obtain a supplemented single-label image set; performing label supplementation on the supplemented single-label image set to obtain a final supplemented image set; and establishing a target multi-label image set to train an image recognition model by using the target multi-label image set.

Abstract: An adjustable hyperspectral detection chip enhanced by a multi-resonance plasmonic mechanism. The detection chip consists of an array of metal nanonail resonator detection units. Each detection unit (1) comprises: a bottom electrode (2), a semiconductor material layer (3), a spacer layer (4), a nanonail array (5), a control material layer (6), a top electrode (7), a peripheral control signal (8), and a driving circuit (9). The positional relationship from top to bottom is the top electrode (7), the control material layer (6), the nanonail array (5), the spacer layer (4), the semiconductor material layer (3), and the bottom electrode (2). The nanonail array (5) is loaded inside the control material layer (6), and the peripheral control signal (8) and the driving circuit (9) are connected to both sides of the control material layer (6).

Abstract: A method for host-side data classification according to embodiments, includes: writing a data sector directly from a host to a storage device; storing the data sector in the storage device; sending a write completion acknowledgement from the storage device to the host; in the background: classifying the data sector on the host to estimate an expected lifetime of the data sector; assigning a data type index to the data sector based on the expected lifetime of the data sector; and asynchronously sending the data type index for the data sector from the host to the storage device.

Abstract: A neural network training method, apparatus, a storage medium, and a computer device are provided. The method includes: obtaining a training sample set, each training sample including a standard label; inputting the each training sample into a neural network model including n attention networks, the n attention networks respectively mapping the each training sample to n subspaces, each of the n subspaces including a query vector sequence, a key vector sequence, and a value vector sequence; calculating a space difference degree between the n subspaces by using the neural network model; calculating an output similarity degree according to an output of the neural network model and the standard label corresponding to the each training sample; and adjusting a model parameter of the neural network model according to the space difference degree and the output similarity degree until a convergence condition is satisfied to obtain a target neural network model.

Abstract: An unmanned aerial vehicle is described herein. The unmanned aerial vehicle includes a fuselage body, a lift mechanism coupled to the fuselage body, and a depth sensing and obstacle avoidance system coupled to the fuselage body. The depth sensing and obstacle avoidance system includes a platform assembly, a pair of stereovision cameras coupled to a platform assembly, and a motor assembly coupled to the fuselage body and to the platform assembly. The platform assembly includes a support member extending between a first end and an opposite second end along a longitudinal axis. The pair of stereovision cameras includes each stereovision camera positioned at an opposite end of the support member. The motor assembly is configured to rotate the platform assembly with respect to the fuselage body about a rotational axis perpendicular to the longitudinal axis of the platform assembly.

Abstract: The present invention relates to composite materials useful for purifying proteins obtained from biological feedstocks. The composite materials of the invention comprise a porous support having an average pore size of 5 to 500 nm, said porous support being filled with a polymer which is cross-linked, wherein the polymer is selected from polyvinylamines or polyallylamines having a weight average molecular weight (Mw) of 2,000 to 500,000 Da and a hydrolysis degree of the formamide groups of at least 66%, with the proviso that a polyvinylamine having a weight average molecular weight (Mw) of 27,200 Da and a hydrolysis degree of 70% and a polyvinylamine having a weight average molecular weight (Mw) of 50,000 Da and a hydrolysis degree of 95% are excluded.

Abstract: The present invention provides a surface plasmon-optical-electrical hybrid conduction nano heterostructure and a preparation method therefor. The structure includes an exciting light source, a semiconductor nano-structure array, a two-dimensionalplasmonic micro-nano structure, a sub-wavelength plasmon polariton guided wave, an emergent optical wave, a one-dimensionalplasmonic micro-nano structure, a wire, a metal electrode, a conductive substrate, a probe molecule, an atomic-force microscopic conductive probe and a voltage source. The method achieves a semiconductor seed crystal with controllable distribution and density by controlling free metal ions, air, water or oxygen on a metal substrate to achieve highly uniform control of the seed crystal, and then strictly controls a length-to-diameter ratio and distribution of a semiconductor structure by continuous growth. Therefore, a new nano optics platform is provided for studying various novel effects produced by interaction between light and substances.

Abstract: A controllable preparation method for a plasmonic nanonail structure is provided. A size of a nanomaterial can be controlled at sub-wavelength. The nanomaterial has good localized surface plasmon resonance effect, and the optical, electrical and mechanical properties of the nanometer material all can be regulated. The plasmonic nanonail is composed of two parts, i.e., a silver nanorod, a gold nanorod or a silver-gold-silver alloy nanorod and an approximate equilateral triangular nano-silver plate growing on the nanorod. A length of the nanorod is controlled within 20-30 nanometers, a diameter of the nanorod is controlled within 10-200 nanometers, a side length of the triangular nano-silver plate is controlled within 20 nanometers to 2 microns, and a size of the triangular plate is less than or equal to the length of the nanorod.

Abstract: A surface plasmon infrared nano-pulse laser having a multi-resonance competition mechanism, consisting of the four parts of a surface plasmon nano-pin resonance chamber (1), a spacer layer (2), a gain medium (3), and a two-dimensional material layer (4). The surface plasmon nano-pin resonance chamber (1) consists of a metal nano rod (11) and one or more nano sheets (12) grown thereon, the surface plasmon nano-pin resonance chamber (1) and the gain medium (3) being isolated by the isolating layer (2), and the two-dimensional material layer (4) covering a surface of the surface plasmon nano-pulse laser; positive and negative electrodes (5) are located at two ends of the surface plasmon nano-pulse laser, and a layer of a two-dimensional material having a feature of saturatable absorption is introduced to a surface of the nano-pin resonance chamber.