Abstract: In one embodiment, an exemplary computer-implemented method of storing point cloud data in an autonomous driving vehicle can include the operations of receiving raw point cloud data from a LiDAR sensor mounted on the autonomous driving vehicle, the raw point cloud data representing cloud data points acquired in response to laser beams emitted at a given angle; retrieving configuration information of the LiDAR sensor, the configuration information including at least a number of laser lines of the LiDAR sensor. The method further includes the operations of constructing, based on the configuration information, a data structure that includes a data entry for each of the cloud data points, the data entry including multiple fields for storing attributes of the cloud data point, each field having a bit width determined based on the configuration information using a predetermined algorithm; and writing the cloud data points to a storage medium using the data structure.

Abstract: A communication system, method, and device for a miniature intelligent sensor. The communication system includes a sensor, a power gateway, and a power Internet of things (IoT) server. The sensor is configured to obtain sensor data and send the sensor data to the power gateway according to a Bluetooth communication protocol. The power gateway is configured to receive the sensor data and perform protocol conversion on the sensor data to obtain protocol conversion data adapted to a message queuing telemetry transport (MQTT) protocol. The power gateway is configured to send the protocol conversion data to the power IoT server according to the MQTT protocol. The power IoT server is configured to receive the protocol conversion data and obtain server receiving data of the power IoT server according to the protocol conversion data.

Abstract: Embodiments relate to systems and methods to optimize quantization of tensors of an AI model. According to one embodiment, a system receives an AI model having one or more layers. The system receives a number of input data for offline inferencing and applies offline inferencing to the AI model based on the input data to generate offline data distributions for the AI model. The system quantizes one or more tensors of the AI model based on the offline data distributions to generate a low-bit representation AI model, where each layer of the AI model includes the one or more tensors, where the one or more tensors include the one or more tensors. In one embodiment, the system applies online inferencing using the low-bit representation AI model to generate online data distributions for a feature map, and quantizes a feature map tensor based on the online data distributions.

Abstract: Batch normalization (BN) layer fusion and quantization method for model inference in artificial intelligence (AI) network engine are disclosed. A method for a neural network (NN) includes merging batch normalization (BN) layer parameters with NN layer parameters and computing merged BN layer and NN layer functions using the merged BN and NN layer parameters. A rectified linear unit (RELU) function can be merged with the BN and NN layer functions.

Abstract: Provided is a polypeptide tag. The amino acid sequence of the polypeptide tag is Xaa1Xaa2Xaa3PHDYNXaa4Xaa5Xaa6 (SEQ ID NO: 37), wherein in the formula, Xaa1, Xaa2, Xaa3, Xaa4, Xaa5, and Xaa6 are each independently an amino acid or none. The polypeptide tag is used for labeling a target protein. In a second aspect, provided is a polypeptide fusion protein, comprising the following two structures: (1) any polypeptide tag according to the first aspect, and (2) a target protein connected to the polypeptide tag. Also provided are an in vitro cell-free protein synthesis system and an application thereof in in vitro protein synthesis. By constructing the polypeptide tag and a target protein as a fusion protein, the expression of the labeled target protein can be effectively increased without removing the polypeptide tag.

Abstract: A tandem DNA element capable of enhancing protein synthesis efficiency, in particular, the nucleic acid construct is formed by an IRES enhancer (such as ScBOI1, ScFLO8, ScNCE102, ScMSN1, KlFLO8, KlNCE102, KlMSN1, KlBOI1) derived from eukaryotic cells (such as yeast), a ? sequence, and a yeast-specific Kozak sequence in tandem. The use of the nucleic acid construct in a yeast-based in vitro biosynthesis system (such as a yeast-based in vitro protein synthesis system) can significantly improve protein synthesis efficiency.

Abstract: What provided is a method for regulating in vitro biosynthesis activity by knocking-out of a nuclease system, comprising: screening five nucleases among numerous nucleases, and performing down-regulation or knocking-out on one of the five nucleases (e.g., EXN53). The method can improve the stability of nucleic acid and the protein production efficiency of an in vitro protein synthesis system.

Abstract: Memory layout and conversion are disclosed to improve neural network (NN) inference performance. For one example, a NN selects a memory layout for a neural network (NN) among a plurality of different memory layouts based on thresholds derived from performance simulations of the NN. The NN stores multi-dimensional NN kernel computation data using the selected memory layout during NN inference. The memory layouts to be selected can be a channel, height, width, and batches (CHWN) layout, a batches, height, width and channel (NHWC) layout, and a batches, channel, height and width (NCHW) layout. If the multi-dimensional NN kernel computation data is not in the selected memory layout, the NN transforms the multi-dimensional NN kernel computation data for the selected memory layout.

Abstract: Provided is biological magnetic microsphere, comprising magnetic microsphere body. Magnetic microsphere body comprises, on outer surface thereof, at least one polymer having linear backbone and side chain. End of linear backbone fixes to outer surface of magnetic microsphere body, other ends of polymer unattach to outer surface of magnetic microsphere body. Biotin links to terminal end of side chain of polymer of biological magnetic microsphere. Further provided are modification to, preparation method for, use of biological magnetic microsphere. Biological magnetic microsphere can be handled and used conveniently, rapidly disperse and rapidly precipitate in solution without need to use large-scale experimental equipment such as high-speed centrifuge, can be connected via biotin with purification element having selectivity (such as avidin, affinity protein, polypeptide/protein tag, etc.

Abstract: The present disclosure provides an additive for an electrolyte for a lithium battery, an electrolyte, and an electrochemical device. The additive includes a compound represented by the following formula I, wherein R1, R2, R3, and R4 are identical to or different from each other, and are independently selected from a group consisting of hydrogen, halogen, substituted or unsubstituted C1-C20 alkyl, and substituted or unsubstituted C6-C18 aryl, and a substituent of the substituted or unsubstituted C1-C20 alkyl or the substituted or unsubstituted C6-C18 aryl can be halogen, C1-C20 alkyl, or C6-C18 aryl; wherein denotes a single bond or a double bond; and wherein n is 1 or 2.

Abstract: Provided are a connection member, a battery and a battery pack. The connection member is used for electrically connecting a plurality of tabs of a current collector and includes a connection plate and at least one protrusion arranged on the connection plate and protruding from one side surface of the connection plate in a thickness direction of the connection plate. At least one end of the at least one protrusion in a length direction of the at least one protrusion is formed as a tip in a plane perpendicular to the thickness direction of the connection plate. The tip is configured to puncture the plurality of tabs in the length direction of the at least one protrusion.

Abstract: A method and system is disclosed for protecting neural network models by segmenting partitions of the models into segments of pre-configured memory size, hashing the segmented models, and concatenating the hash segments. The concatenated hash segment may be further hashed, encrypted, and stored with the neural network models as an executable loadable file (ELF) in memories external to the neural network prior to the use of the models by the neural network. The models may include model weights of the inference layers and metadata. The model weights and the metadata may be hashed as separate hash segments and concatenated. Segmenting the models into segments of pre-configured memory size and hashing the segmented models offline prior to the operation of the neural network enables rapid validation of the models when the models are used in the inference layers during online operation of the neural network.

Abstract: Embodiments of the invention disclose an integrated circuit and a method for improving utilization of multiply and accumulate (MAC) units on the integrated circuit in an artificial intelligence (AI) engine. In one embodiment, the integrated circuit can include a scheduler for allocating the MAC units to execute a neural network model deployed on the AI engine to process input data. The scheduler includes status information for the MAC units, and can select one or more idle MAC units based on the status information for use to process the feature map slice. The integrated circuit can dynamically map idle MAC units to an input feature map, thereby improving utilization of the MAC units. A pair of linked list, each with a reference head, can be provided in a static random access memory (SRAM) to store only feature map slices and weights for a layer that is currently being processed.

Abstract: The present disclosure discloses a current collector assembly, a battery cell, and a battery pack. The current collector assembly includes a current collector and a conductive connection assembly, the current collector includes a plurality of tabs stacked in a thickness direction of the tab, the conductive connection assembly includes a protection plate and a connection plate, the protection plate is arranged on a side of the plurality of tabs in the thickness direction, the connection plate is arranged on the other side of the plurality of tabs in the thickness direction, a partial region of one of the connection plate and the protection plate is deformed to form a protrusion which, during the forming, pierces through the plurality of tabs to be electrically connected to the plurality of tabs.

Abstract: Provided are a preparation method of a current collector assembly, a current collector assembly, a battery cell, and a battery pack. The preparation method of the current collector assembly includes: abutting the at least one of the connection plate or the protection plate having the protrusion against the plurality of tabs; generating a high-frequency vibration between the at least one of the connection plate or the protection plate having the protrusion and the plurality of tabs; and piercing the plurality of tabs sequentially by the protrusion through the high-frequency vibration. A vibration frequency of the high-frequency vibration ranges from 15 KHz to 40 KHz.

Abstract: A biomagnetic microsphere and a preparation method and a method for protein isolation and purification therefor. The outer surface of a magnetic microsphere body of the biomagnetic microsphere has at least one liner polymer with a branched chain; one end of the linear polymer with a branched chain is covalently coupled to the outer surface of the magnetic microsphere body, and other parts are free on the outer surface of the magnetic microsphere body; a backbone of the linear polymer is a polyolefin backbone, and no cross-linking agent is required in the backbone forming process of the linear polymer. The prepared biomagnetic microsphere can implement efficient elution of target proteins and effectively reduce the retention time and retention ratio of the target proteins, and it is easy to operate and widely used.

Abstract: A method for synthesizing and application embedded alkaline earth metal oxide solid alkali includes: firstly, synthesizing an alkaline earth metal organic skeleton with single or multiple alkaline earth metals (Mg, Ca and Sr) as central metal elements; and then controlling the heating process to carry out high-temperature pyrolysis in a non-oxidizing atmosphere, so that the alkaline earth metal oxide are embedded in the nano carbon sheet to obtain a solid alkali catalyst. Finally, the catalyst is used to catalyze the transesterification of palm oil and methanol to produce biodiesel. The active site of the solid alkali obtained by the method is anchored on the nano-like carbon sheet, so that the active site is directly exposed on the surface of the catalyst, the catalytic activity is improved, the loss of the active site is inhibited, and the stability of the solid alkali catalyst is enhanced.

Abstract: Provided is a method for quantitatively co-expressing multiple proteins in an in vitro cell-free protein synthesis system, comprising the following steps: (1) establishing a standard curve of the relationship between standard protein concentration and luminescence value; (2) creating a vector containing a target protein gene, and obtaining an in vitro protein synthesis system; (3) establishing a curve of the relationship between the target protein concentration and the vector concentration; (4) calculating the concentration or/and the concentration ratio of the vector quantitatively co-expressing multiple target proteins; (5) quantitatively co-expressing the target proteins.

Abstract: The present disclosure relates to the technical field of fiber-optic gas sensing and laser photoacoustic spectroscopy, and reliability of a gas detection system is improved by actively selecting a working frequency of low noise interference combined with an optical fiber photoacoustic sensing probe capable of isolating high-frequency noise. A gas enters a photoacoustic microcavity through gaps on a sound-sensitive diaphragm after diffusing into a miniature air chamber through a plurality of micropores. Photoacoustic excitation light is incident into the photoacoustic microcavity through a fiber-optic collimator and then excited to generate a photoacoustic pressure wave to cause the sound-sensitive diaphragm to vibrate periodically. An end face of a single-mode optical fiber and the sound-sensitive diaphragm constitutes a fiber-optic Fabry-Perot interferometer. The interferometer measures a deflection of the diaphragm and inverts a concentration of the to-be-measured gas.