Abstract: The embodiments of the present application discloses an image processing method, an image processing apparatus, an image processing device and a computer-readable storage medium.

Abstract: According to an example aspect of the present invention, there is provided an apparatus comprising memory configured to store convolutional artificial neural network information comprising at least one filter definition, and at least one processing core configured to generate, from a preceding layer, a convolutional result of a succeeding layer of the artificial neural network in accordance with the at least one filter definition, and generate, from the convolutional result, an activation result of the succeeding layer by using an activation function, the activation function taking three arguments, the three arguments being derived from the convolutional result.

Abstract: Disclosed is a slag discharging method in a process of producing ultra-low phosphorus steel, which relates to the technical field of iron and steel smelting, and in which molten steel is mixed with lime first to produce basic slag; then converting is performed with oxygen to increase the oxidizability of the basic slag; and a carbon-containing reducing agent is finally added, so that in the process that the carbon is oxidized to release a large amount of carbon monoxide gas, phosphates are captured, and the basic slag is rapidly foamed and overflows from the opening of the steel ladle, so that conditions are no longer available for rephosphorization. Also disclosed is a method for producing ultra-low phosphorus steel, which includes the above-described slag discharging method in a process of producing ultra-low phosphorus steel, and refining and ingotting after slag discharge.

Abstract: According to an example aspect of the present invention, there is provided an apparatus comprising memory configured to store convolutional artificial neural network information comprising at least one filter definition, and at least one processing core configured to generate, from a preceding layer, a convolutional result of a succeeding layer of the artificial neural network in accordance with the at least one filter definition, and generate, from the convolutional result, an activation result of the succeeding layer by using an activation function, the activation function taking three arguments, the three arguments being derived from the convolutional result.

Abstract: Methods and apparatus, including computer program products, are provided for machine learning including deep convolutional neural networks. In some example embodiments, there may be provided a method that includes receiving, at a trained machine learning model, a portion of a test image; activating, at the machine learning model, a convolutional result formed based on the portion of the test image, the activation based on neighboring regions in the test image; and providing, by the machine learning model, an activated output. Related systems, methods, and articles of manufacture are also described.

Abstract: Embodiments of the present invention disclose a binocular image matching method, apparatus, device, and storage medium. The method comprises: performing target detection on a first image to obtain a first bounding box of a target in the first image; determining a second bounding box corresponding to the first bounding box in a second image; and obtaining a third bounding box of the target in the second image by regressing the second bounding box. The technical solutions realize accurate matching of targets in a binocular image without requiring performing target detection on both images in the binocular image, and then use a matching algorithm to match the targets detected in the two images, thereby greatly reducing the calculation overhead of target matching in the binocular image.

Abstract: An information synchronization method, a terminal and a system, which relates to the technical field of communication. The method includes: by the first terminal, sending a controlling request of a target terminal to the controlling server; by the first terminal, receiving a state feedback information sent by the first terminal, according to the state feedback information, updating an information of the first terminal; and by the first terminal, sending the state feedback information to each of the at least one second terminal, so that the second terminal updates an information of the second terminal according to the state feedback information.

Abstract: A flavor system for a non-animal derived protein consumable is provided. The flavor system includes an aqueous component; a protein binder including a mixture of at least one terpene and at least one carbonyl compound; one or more off-note blocking compounds; and a flavorant.

Abstract: An object detection method, a neural network training method, an apparatus, and an electronic device include: obtaining, through prediction, multiple fused feature graphs from images to be processed, through a deep convolution neural network for target region frame detection, obtaining multiple first feature graphs from a first subnet having at least one lower sampling layer, obtaining multiple second feature graphs from a second subnet having at least one upper sampling layer, and obtaining fused graph by fusing multiple first feature graphs and multiple second feature graphs respectively; and obtaining target region frame data according to the multiple fused feature graphs. Because the fused feature graphs better represent semantic features on high levels and detail features on low levels in images, target region frame data of big and small objects in images can be effectively extracted according to the fused feature graphs, thereby improving accuracy and robustness of object detection.

Abstract: In one embodiment, a method includes accessing posts in a social-networking system. Each of the posts is unlabeled with respect to whether the post is known to be spam. The method also includes determining a posting user who submitted the post to the social-networking system and a recipient user to whom the post is addressed. The method further includes determining a first vector representation of the posting user and a second vector representation of the recipient user based on one or more features associated with the post, the posting user, and the recipient user. The method still further includes comparing the vector representations and building a machine learning model for automatically detecting spam posts in the social-networking system using a subset of the plurality of posts as non-spam training data.

Abstract: In one embodiment, a method includes accessing posts in a social-networking system. Each of the posts is unlabeled with respect to whether the post is known to be spam. The method also includes determining a posting user who submitted the post to the social-networking system and a recipient user to whom the post is addressed. The method further includes determining a first vector representation of the posting user and a second vector representation of the recipient user based on one or more features associated with the post, the posting user, and the recipient user. The method still further includes comparing the vector representations and building a machine learning model for automatically detecting spam posts in the social-networking system using a subset of the plurality of posts as non-spam training data.

Abstract: Compositions and methods of preparing energy storage device electrode active materials and electrodes are described. A two-step synthesis process may be utilized to prepare single crystal electrode active materials and electrodes, such as a single crystal nickel-cobalt-aluminum material. In some embodiments, the two step synthesis process includes a first and a second lithiation step.

Abstract: Disclosed is a slag discharging method in a process of producing ultra-low phosphorus steel, which relates to the technical field of iron and steel smelting, and in which molten steel is mixed with lime first to produce basic slag; then converting is performed with oxygen to increase the oxidizability of the basic slag; and a carbon-containing reducing agent is finally added, so that in the process that the carbon is oxidized to release a large amount of carbon monoxide gas, phosphates are captured, and the basic slag is rapidly foamed and overflows from the opening of the steel ladle, so that conditions are no longer available for rephosphorization. The slag discharging method is simple and convenient to operate, does not have high requirements on the equipment, has relatively good dephosphorization effect, and can be used to prepare an ultra-low phosphorus steel containing less than 0.003% phosphorus.

Abstract: A control method for a series hybrid power system or synergic electric power supply and a control device thereof are disclosed. The control device includes: a generator reference current value generating module and a generator current control module. The generator reference current value generating module is configured to generate a reference current value of a generator according to an actual current value of an energy storage device, a reference current value of the energy storage device, an actual voltage value of a DC bus, and a reference voltage value of the DC bus. The generator current control module is configured to generate a control signal of a rectifier according to the reference current value of the generator, the actual value of any two-phase current of the generator, and an actual angular velocity of the generator, thereby controlling the series hybrid power system or the synergic electric power supply.

Abstract: A flavor system for a non-animal derived protein consumable is provided. The flavor system includes an aqueous component; a protein binder including a mixture of at least one terpene and at least one carbonyl compound; one or more off-note blocking compounds; and a flavorant.

Abstract: According to an example aspect of the present invention, there is provided an apparatus comprising memory configured to store convolutional artificial neural network information comprising at least one filter definition, and at least one processing core configured to generate, from a preceding layer, a convolutional result of a succeeding layer of the artificial neural network in accordance with the at least one filter definition, and generate, from the convolutional result, an activation result of the succeeding layer by using an activation function, the activation function taking three arguments, the three arguments being derived from the convolutional result.

Abstract: A control method for a series hybrid power system or synergic electric power supply and a control device thereof are disclosed. The control device includes: a generator reference current value generating module and a generator current control module. The generator reference current value generating module is configured to generate a reference current value of a generator according to an actual current value of an energy storage device, a reference current value of the energy storage device, an actual voltage value of a DC bus, and a reference voltage value of the DC bus. The generator current control module is configured to generate a control signal of a rectifier according to the reference current value of the generator, the actual value of any two-phase current of the generator, and an actual angular velocity of the generator, thereby controlling the series hybrid power system or the synergic electric power supply.

Abstract: Methods of preparing electrodes for use in rechargeable battery using two lithiation steps wherein, including a first lithiation step conducted at higher temperatures than the second lithiation step.

Abstract: Methods and apparatus, including computer program products, are provided for machine learning including deep convolutional neural networks. In some example embodiments, there may be provided a method that includes receiving, at a trained machine learning model, a portion of a test image; activating, at the machine learning model, a convolutional result formed based on the portion of the test image, the activation based on neighboring regions in the test image; and providing, by the machine learning model, an activated output. Related systems, methods, and articles of manufacture are also described.