Abstract: A medical image analysis method includes: reading an original medical image; performing image classification and object detection on the original medical image to generate a first classification result and a plurality of object detection results by a plurality of complementary artificial intelligence (AI) models; performing object feature integration and transformation on a first detection result and a second detection result among the object detection results to generate a transformation result by a features integration and transformation module; and performing machine learning on the first classification result and the transformation result to generate an image interpretation result by a machine learning module and display the image interpretation result.

Abstract: A method for continuous measurement of river flow based on satellite big data is provided. The method includes: determining a river reach to conduct flow measurement, reconstructing a cross section of a river channel based on satellite big data, calculating real-time water levels by coupling data of various types of satellites, and performing flow calculation and compilation. The method solves the difficult problem of river flow measurement or continuous measurement of river flow in uninhabited areas, fills the blank of satellite-based flow measurement according to the principle of river dynamics, and greatly expands the range of river flow measurement.

Abstract: A method for continuous measurement of river flow based on satellite big data is provided. The method includes: determining a river reach to conduct flow measurement, reconstructing a cross section of a river channel based on satellite big data, calculating real-time water levels by coupling data of various types of satellites, and performing flow calculation and compilation. The method solves the difficult problem of river flow measurement or continuous measurement of river flow in uninhabited areas, fills the blank of satellite-based flow measurement according to the principle of river dynamics, and greatly expands the range of river flow measurement.

Abstract: Methods, systems, computer-readable media, and apparatuses are presented for computer-assisted visualization of network devices. One example involves receiving a plurality of standardized network description files describing a plurality of vehicular communication networks connecting a plurality of electronic control units (ECU) for a vehicle. Each of the plurality of standardized network description files may describe a vehicular communication network in the plurality of vehicular communication networks. Each vehicular communication network may comprise a subset of the plurality of ECUs and one or more network communications paths interconnecting the subset of ECUs. The example can further involve automatically generating, based on the standardized network description files, a visual topology representation of the plurality of vehicular communication networks connecting to the plurality of ECUs.

Abstract: A medical image analysis method includes: reading an original medical image; performing image classification and object detection on the original medical image to generate a first classification result and a plurality of object detection results by a plurality of complementary artificial intelligence (AI) models; performing object feature integration and transformation on a first detection result and a second detection result among the object detection results to generate a transformation result by a features integration and transformation module; and performing machine learning on the first classification result and the transformation result to generate an image interpretation result by a machine learning module and display the image interpretation result.

Abstract: A training system and method of object detection model is disclosed. The training system includes an object detection model and a loss calculation module. The object detection model is configured to generate an output image according to an input image. The loss calculation module, coupled to the object detection model, is configured to calculate a total classification loss value according to the output image and a solution image, calculate a loss value according to the total classification loss value, and transmit the loss value to the object detection model. The total classification loss value is calculated according to a number of classification losses corresponding to a number of object types. Each classification loss corresponding to each object type is calculated according to a first parameter, a second parameter and a third parameter.

Abstract: A classification device and a classification method based on a neural network are provided. A heterogeneous integration module includes a convolutional layer, a data normalization layer, a connected layer and a classification layer. The convolutional layer generates a first feature map according to a first image data. The data normalization layer normalizes a first numerical data to generate a first normalized numerical data. The first numerical data corresponds to the first image data. The connected layer generates a first feature vector according to the first feature map and the first normalized numerical data. The classification layer generates a first classification result corresponding to a first time point according to the first feature vector. The heterogeneous integration module generates a second classification result corresponding to a second time point.

Abstract: Certain embodiments are described that provide a method and computer readable media for testing battery cells. A unique identifier (e.g., barcode) is affixed to a battery cell which allows it to be tracked across separate tests as a cell, in a module, string, pack, etc. Using a GUI, the unique identifier is recorded in a database along with at least a battery cell manufacturer and a battery cell model. A designation of the particular tester channel or module or string location is entered into the database in association with the unique identifier. Test results of the first test are electronically transferred from the first tester to the database along with the corresponding channel designations.

Abstract: A board defect filtering method is provided. The method includes: receiving a defect list; obtaining a plurality of defect images of a plurality of defect records on the defect list; receiving a circuit layout image; analyzing a defect location of a first defect image of the plurality of defect images according to the circuit layout image; cropping the first defect image to obtain a first cropped defect image according to the defect location; inputting the first cropping defect image to a defect classifying model; and determining whether the first defect image is a qualified product image or not according to an output result of the defect classifying model.

Abstract: Certain embodiments are described that provide a method and computer readable media for testing battery cells. A unique identifier (e.g., barcode) is affixed to a battery cell which allows it to be tracked across separate tests as a cell, in a module, string, pack, etc. Using a GUI, the unique identifier is recorded in a database along with at least a battery cell manufacturer and a battery cell model. A designation of the particular tester channel or module or string location is entered into the database in association with the unique identifier. Test results of the first test are electronically transferred from the first tester to the database along with the corresponding channel designations.

Abstract: A training system and method of object detection model is disclosed. The training system includes an object detection model and a loss calculation module. The object detection model is configured to generate an output image according to an input image. The loss calculation module, coupled to the object detection model, is configured to calculate a total classification loss value according to the output image and a solution image, calculate a loss value according to the total classification loss value, and transmit the loss value to the object detection model. The total classification loss value is calculated according to a number of classification losses corresponding to a number of object types. Each classification loss corresponding to each object type is calculated according to a first parameter, a second parameter and a third parameter.

Abstract: A neural network model fusion method and an electronic device using the same are provided. The neural network model fusion method includes the following steps. An image is received. The image is analyzed through several neural network models. The neural network models include at least two of a degree classification model, a multi-class identification model and an object detection model. Several analysis results are obtained according to the neural network models. These analysis results are converted into a number of conversion factors. The conversion factors are inputted into a fusion model to obtain a fusion result.

Abstract: Certain embodiments are described that provide a method and computer readable media for testing battery cells. A unique identifier (e.g., barcode) is affixed to a battery cell which allows it to be tracked across separate tests as a cell, in a module, string, pack, etc. Using a GUI, the unique identifier is recorded in a database along with at least a battery cell manufacturer and a battery cell model. A designation of the particular tester channel or module or string location is entered into the database in association with the unique identifier. Test results of the first test are electronically transferred from the first tester to the database along with the corresponding channel designations.

Abstract: A neural network model fusion method and an electronic device using the same are provided. The neural network model fusion method includes the following steps. An image is received. The image is analyzed through several neural network models. The neural network models include at least two of a degree classification model, a multi-class identification model and an object detection model. Several analysis results are obtained according to the neural network models. These analysis results are converted into a number of conversion factors. The conversion factors are inputted into a fusion model to obtain a fusion result.

Abstract: A board defect filtering method is provided. The method includes: receiving a defect list; obtaining a plurality of defect images of a plurality of defect records on the defect list; receiving a circuit layout image; analyzing a defect location of a first defect image of the plurality of defect images according to the circuit layout image; cropping the first defect image to obtain a first cropped defect image according to the defect location; inputting the first cropping defect image to a defect classifying model; and determining whether the first defect image is a qualified product image or not according to an output result of the defect classifying model.

Abstract: Certain embodiments are described that provide a method and computer readable media for testing battery cells. A unique identifier (e.g., barcode) is affixed to a battery cell which allows it to be tracked across separate tests as a cell, in a module, string, pack, etc. Using a GUI, the unique identifier is recorded in a database along with at least a battery cell manufacturer and a battery cell model. A designation of the particular tester channel or module or string location is entered into the database in association with the unique identifier. Test results of the first test are electronically transferred from the first tester to the database along with the corresponding channel designations.

Abstract: Methods, systems, computer-readable media, and apparatuses are presented for computer-assisted visualization of network devices. One example involves receiving a plurality of standardized network description files describing a plurality of vehicular communication networks connecting a plurality of electronic control units (ECU) for a vehicle. Each of the plurality of standardized network description files may describe a vehicular communication network in the plurality of vehicular communication networks. Each vehicular communication network may comprise a subset of the plurality of ECUs and one or more network communications paths interconnecting the subset of ECUs. The example can further involve automatically generating, based on the standardized network description files, a visual topology representation of the plurality of vehicular communication networks connecting to the plurality of ECUs.

Abstract: A thermostatic control LED thermal module includes at least one fan, a heat dissipation unit, a sensing unit and a controller. One side of the heat dissipation unit is mated with the fan, while the other side of the heat dissipation unit is attached to an LED unit. The sensing unit is electrically connected to the LED unit for detecting the temperature of the LED unit and generating a sensing signal. The controller is electrically connected to the fan and the sensing unit. According to the received sensing signal, the controller operates and processes to generate a control signal to control rotational speed of the fan so as to achieve an excellent heat dissipation effect.

Abstract: A rotatable heat dissipating device includes a heat-transfer element and a heat-dissipation element. The heat-transfer element includes a flat main body, a heat transfer section extended from one face of the main body. The heat-dissipation element includes base having a recess provided on one face oriented toward the main body, a heat radiating section, and a connecting section extended between the base and the radiating section. The main body is rotatably received in and connected to the recess of the base, enabling the rotatable heat dissipating device to flexibly connect to a heat-producing unit in different directions without the need of using any external adaptor. Therefore, the rotatable heat dissipating device provides upgraded heat dissipation efficiency and can avoid thermal resistance due to too many adaptors between different elements.