Abstract: An ultrasonic atomization type electronic cigarette includes an ultrasonic atomizer having an ultrasonic atomization sheet, an electronic cigarette housing, an upper cover assembly, and a locking structure. The angle ? between the ultrasonic atomization sheet and the central axis of the electronic cigarette is 5° to 85°, preferably 15° to 55°. The housing is provided with an air inlet groove. A first support in the housing includes a receiving cavity for receiving the ultrasonic atomizer. The upper cover assembly may be used for opening the receiving cavity, and the locking structure may be used for locking the upper cover assembly. A liquid guide structure in the atomizer may include a liquid guide body in which a side of a first liquid guide layer is sandwiched between a side of a second liquid guide layer and an ultrasonic atomization surface.

Abstract: An ultrasonic electronic cigarette atomizing core and atomizer is provided, the atomizing core including an atomizing core sleeve, and tobacco tar guide cotton and an atomizing piece in the atomizing core sleeve.

Abstract: A method may include determining, with a queue availability module, that an entry is available in a queue, asserting a bit in a register based on determining that an entry is available in the queue, determining, with a processor, that the bit is asserted, and processing, with the processor, the entry in the queue based on determining that the bit is asserted. The method may further include storing the register in a tightly coupled memory associated with the processor. The method may further include storing the queue in the tightly coupled memory. The method may further include determining, with the queue availability module, that an entry is available in a second queue, and asserting a second bit in the register based on determining that an entry is available in the second queue. The method may further include finding the first bit in the register using a find first instruction.

Abstract: A method and apparatus of open set recognition, and a computer-readable storage medium are disclosed. The method comprises acquiring auxiliary data and training data of known categories for open set recognition, training a neural network alternately using the auxiliary data and the training data, until convergence; extracting a feature of data to be recognized for open set recognition, using the trained neural network; and recognizing a category of data to be recognized, based on the feature of the data to be recognized.

Abstract: An ultrasonic electronic cigarette is disclosed. The electronic cigarette includes an atomizer (1) and a battery assembly (2) detachably connected to each other. The atomizer (1) comprises an ultrasonic atomization sheet (3), and the battery assembly (2) comprises a first battery electrode (4) and a second battery electrode (5) for supplying power to the atomizer (1). When the atomizer (1) is connected to the battery assembly (2), a first pole of the ultrasonic atomization sheet (3) is electrically connected to the first battery electrode (4), and a second pole of the ultrasonic atomization sheet (3) is electrically connected to the second battery electrode (5).

Abstract: The present invention discloses an oscillation control circuit for an ultrasonic atomization sheet, and an ultrasonic electronic cigarette. The oscillation control circuit includes a DC boost module, a separately excited excitation module, and a microcontroller. A power module is connected to the ultrasonic atomization sheet through the DC boost module and the separately excited excitation module, and the separately excited excitation module is electrically connected to the microcontroller.

Abstract: Disclosed is an atomizer, comprising an atomizer body, a suction nozzle assembly, an air inlet, an atomization sheet, a tobacco tar chamber, a tobacco tar guide device and an air outflow passage, in which a tobacco tar separation chamber is provided in the middle of the tobacco tar chamber and an accommodating chamber is provided in the tobacco tar separation chamber.

Abstract: An ultrasonic electronic cigarette atomizer is disclosed, including an atomization core, an atomizer housing, and a liquid bin. When liquid is filled into the liquid bin, the liquid passing hole is closed and the air outlet channel is closed, to avoid leakage or immersion of liquid, the atomizer is clean and sanitary, waste is avoided, and cost is saved. The whole operation is very simple, the atomization core is convenient to replace, hands are not stained with liquid, and the liquid is convenient to fill. The cigarette liquid is unlikely to flow out through the liquid filling hole, and large granular cigarette liquid generated in atomization can be recycled.

Abstract: Disclosed are a method and apparatus for training a classification model and a method and apparatus for classifying. A method for classifying comprises: extracting a feature from to-be-tested information inputted to a classification model having been trained; compressing the extracted feature into a low dimensional hidden feature capable of representing the to-be-tested information; performing decompression on the hidden feature to obtain a decompressed feature; performing rebuilding on the to-be-tested information based on the decompressed feature, to obtain reconstructed to-be-tested information; judging, based on a rebuild loss between the to-be-tested information and the reconstructed to-be-tested information, whether the to-be-tested information belongs to a known class or an unknown class; and performing classification on the to-be-tested information, via the classification model having been trained, in a case where it is determined that the to-be-tested information belongs to a known class.

Abstract: Disclosed are a method and apparatus for training a classification model and a method and apparatus for classifying. A method for classifying comprises: extracting a feature from to-be-tested information inputted to a classification model having been trained; compressing the extracted feature into a low dimensional hidden feature capable of representing the to-be-tested information; performing decompression on the hidden feature to obtain a decompressed feature; performing rebuilding on the to-be-tested information based on the decompressed feature, to obtain reconstructed to-be-tested information; judging, based on a rebuild loss between the to-be-tested information and the reconstructed to-be-tested information, whether the to-be-tested information belongs to a known class or an unknown class; and performing classification on the to-be-tested information, via the classification model having been trained, in a case where it is determined that the to-be-tested information belongs to a known class.

Abstract: A method and apparatus of open set recognition, and a computer-readable storage medium are disclosed. The method comprises acquiring auxiliary data and training data of known categories for open set recognition, training a neural network alternately using the auxiliary data and the training data, until convergence; extracting a feature of data to be recognized for open set recognition, using the trained neural network; and recognizing a category of data to be recognized, based on the feature of the data to be recognized.