Abstract: Described herein are systems, methods, and other techniques for training a generative adversarial network (GAN) to perform an image-to-image transformation for recognizing text. A pair of training images are provided to the GAN. The pair of training images include a training image containing a set of characters in handwritten form and a reference training image containing the set of characters in machine-recognizable form. The GAN includes a generator and a discriminator. The generated image is generated using the generator based on the training image. Update data is generated using the discriminator based on the generated image and the reference training image. The GAN is trained by modifying one or both of the generator and the discriminator using the update data.

Abstract: An inertial measurement unit (IMU) device includes a circuit board assembly including a rigid circuit board and a flexible circuit board, an IMU sensor disposed on the rigid circuit board, and a heat preservation system. The IMU sensor is electrically connected to an external element through the flexible circuit board to transmit at least one of a signal or power between the IMU sensor and the external element. The heat preservation system includes a heat source, a heat preservation body with a receiving space to accommodate the IMU sensor, and a heat conductive member configured to transfer heat from the heat source to the IMU sensor to maintain the IMU sensor at a preset temperature. The heat conductive member includes an electrically insulating and thermally conductive silicone.

Abstract: A computing server may continuously update a set of nodes that are addable to a data tree based on past interactions of the user with one or more nodes. The computing server may track a recently interacted set of interacted nodes with which the user has interacted within a number of past interactions. The computing server may select a pool of candidate nodes based on the recently interacted set. At least one of the candidate nodes is within a domain boundary of one of the interacted nodes that is in the recently interacted set. The domain boundary may be determined by the degree of relationship. The computing server may present one or more candidate nodes in the pool as a version of the continuously updated set of nodes. The computing server may update the pool of candidate nodes as additional interactions performed by the user updates the recently interacted set.

Abstract: An aerial vehicle control method includes detecting a flight control instruction during returning of an aerial vehicle to a return point along a return trajectory according to an auto return instruction. The flight control instruction determines a predetermined trajectory different from the return trajectory. The method further includes generating a superimposed instruction by superimposing return point position information indicating the return point and the flight control instruction. The superimposed instruction determines a flight trajectory that is an integration of the return trajectory and the predetermined trajectory and that is different from the return trajectory and the predetermined trajectory. The method also includes controlling the aerial vehicle to fly along the flight trajectory according to the superimposed instruction.

Abstract: A substrate container system comprises a container body having a bottom face, a front opening that enables passage of a substrate, and a back opening opposing the front opening, the back opening having a width smaller than that of the front opening; and a back cover that covers the back opening and establishes sealing engagement with the container body, wherein the back cover comprises a first gas inlet structure that bendingly extends under the bottom face of the container body upon assembly; wherein the first gas inlet structure comprises a downward facing gas intake port opposing the bottom face of the container body.

Abstract: Provided are methods of preparing compounds and pharmaceutical compositions for treating Filoviridae virus infections The compounds, compositions, and methods provided are particularly useful for the treatment of Marburg virus, Ebola virus and Cueva virus infections.

Abstract: Systems, methods, and devices are provided for controlling an unmanned aerial vehicle (UAV) associated with flight response measures. The flight response measure may be generated by assessing one or more flight-restriction strips, assessing at least one of a location or a movement characteristic of the UAV relative to the one or more flight-restriction strips, and directing, with aid of one or more processors, the UAV to take one or more flight response measures based on at least one of the location or movement characteristic of the UAV relative to the one or more flight-restriction strips.

Abstract: An audio processing method is configured to upsample a first head related transfer function (HRTF) into a second HRTF. The first HRTF defines first audio feature values distributed at first intersection nodes in a spherical coordinate system over audio frequencies. The first intersection nodes are arranged with a first spatial resolution in the spherical coordinate system. The first HRTF is upsampled into the second HRTF by a spherical convolutional neural network model. The second HRTF defines a plurality of second audio feature values distributed at second intersection nodes in the spherical coordinate system over the audio frequencies. The second intersection nodes in the second HRTF are arranged with a second spatial resolution higher than the first spatial resolution of the first intersection nodes in the first HRTF.

Abstract: Embodiments described herein relate generally to a methodology of efficient object classification within a visual medium. The methodology utilizes a first neural network to perform an attention based object localization within a visual medium to generate a visual mask. The visual mask is applied to the visual medium to generate a masked visual medium. The masked visual medium may be then fed into a second neural network to detect and classify objects within the visual medium.

Abstract: Systems, methods, and devices are provided for controlling an unmanned aerial vehicle (UAV) associated with flight response measures. The flight response measure may be generated by assessing one or more flight-restriction strips, assessing at least one of a location or a movement characteristic of the UAV relative to the one or more flight-restriction strips, and directing, with aid of one or more processors, the UAV to take one or more flight response measures based on at least one of the location or movement characteristic of the UAV relative to the one or more flight-restriction strips.

Abstract: A maintenance method of a gas detecting device includes the following steps: providing the gas detecting device that includes a shell component, a sensing component, a dust blocking element, and a detecting area, the sensing component including a sensor, a reference light source, and a processor, and the dust blocking element including a cover; the reference light source generating a reference beam projecting toward the cover and the detecting area, and the reference beam generating a feedback beam projecting toward the cover and the sensor through being reflected by a part of the shell component positioned at the detecting area; the sensor receiving the feedback beam to generate a measured information; and the processor comparing the measured information to an initial information in the processor, so as to obtain a pollution degree information of the cover.

Abstract: Described herein is a method for preparing a dehydrated and de-alcoholized flavor composition by decreasing or removing water and ethanol from a flavor composition, the method including the steps of treating the flavor composition by a dehydration process and dealcoholization process. Also described herein are flavor compositions obtainable by this method, flavored consumer products comprising the same, and methods and uses thereof.

Abstract: A color filter substrate and a method of manufacturing the same and a liquid crystal display device are disclosed. A display area is provided with a first spacer, and a non-display area is provided with a second spacer. The non-display area is further provided with a virtual color resist layer. Color of the virtual color resist layer is the same as color of the third color resist layer, and a height of the first spacer is greater than a height of the second spacer.

Abstract: Described herein are systems, methods, and other techniques for extracting one or more keyphrases from an input text. The input text may include a plurality of words. A plurality of token-level attention matrices may be generated using a transformer-based machine learning model. The plurality of token-level attention matrices may be converted into a plurality of word-level attention matrices. A set of candidate phrases may be identified from the plurality of words based on the plurality of word-level attention matrices. The one or more keyphrases may be selected from the set of candidate phrases.

Abstract: An inertial measurement unit includes a sensor and a heat preservation system. The heat preservation system includes a heat preservation body and a heat source. The sensor is positioned on the heat preservation body. The heat source is configured to generate heat. The heat preservation body is configured to transfer the heat from the heat source to the sensor to maintain a preset temperature in a space surrounding the sensor.

Abstract: An audio processing method is configured to upsample a first head related transfer function (HRTF) into a second HRTF. The first HRTF defines first audio feature values distributed at first intersection nodes in a spherical coordinate system over audio frequencies. The first intersection nodes are arranged with a first spatial resolution in the spherical coordinate system. The first HRTF is upsampled into the second HRTF by a spherical convolutional neural network model. The second HRTF defines a plurality of second audio feature values distributed at second intersection nodes in the spherical coordinate system over the audio frequencies. The second intersection nodes in the second HRTF are arranged with a second spatial resolution higher than the first spatial resolution of the first intersection nodes in the first HRTF.

Abstract: A simplified handwriting recognition approach includes a first network comprising convolutional neural network comprising one or more convolutional layers and one or more max-pooling layers. The first network receives an input image of handwriting and outputs an embedding based thereon. A second network comprises a network of cascaded convolutional layers including one or more subnetworks configured to receive an embedding of a handwriting image and output one or more character predictions. The subnetworks are configured to downsample and flatten the embedding to a feature map and then a vector before passing the vector to a dense neural network for character prediction. Certain subnetworks are configured to concatenate an input embedding with an upsampled version of the feature map.

Abstract: Provided are methods of preparing compounds and pharmaceutical compositions comprising a compound Formula VIII for treating Filoviridae virus infections. In one aspect, the compound of Formula VIII is formed from a reaction mixture comprising the compound of Formula IX, the compound of Formula X, a coupling agent such as magnesium chloride and a non-nucleophilic base such as diisopropylethylamine. The compound of Formula IX can be formed from a compound of Formula V and a cyanating agent. The compound of Formula V can be synthesized from a reaction mixture comprising a deprotonating agent such as phenylmagnesium chloride; a silylating agent such as chlorotrimethylsilane; a coupling agent such as isopropylmagnesium chloride, an additive such as LaCl3-2LiCl, LaCl3, CeCl3, NdCl3, or YCl3; a compound of Formula VI; and 7-iodopyrrolo[2,1-f][1,2,4]triazin-4-amine. The compounds, compositions, and methods provided are particularly useful for the treatment of Marburg virus, Ebola virus and Cueva virus infections.

Abstract: The present invention is a surface treated drug-loaded solid (e.g., non-porous) microparticle that aggregates in vivo to form a consolidated larger particle for medical therapy. In one embodiment, the particles are used for ocular therapy. Processes for producing the surface treated microparticle and injectable formulations which include the surface treated microparticle are also provided. When used in the eye, long-term consistent intraocular delivery can be achieved without disrupting vision and minimizing undesirable inflammatory responses.