Abstract: A method for scene perception using video captioning based on a spatio-temporal graph model is described. The method includes decomposing the spatio-temporal graph model of a scene in input video into a spatial graph and a temporal graph. The method also includes modeling a two branch framework having an object branch and a scene branch according to the spatial graph and the temporal graph to learn object interactions between the object branch and the scene branch. The method further includes transferring the learned object interactions from the object branch to the scene branch as privileged information. The method also includes captioning the scene by aligning language logits from the object branch and the scene branch according to the learned object interactions.

Abstract: Introduced here are computer programs and associated computer-implemented techniques for deriving insights into the health of patients through analysis of audio data generated by electronic stethoscope systems. A diagnostic platform may be responsible for examining the audio data generated by an electronic stethoscope system so as to gain insights into the health of a patient. The diagnostic platform may employ heuristics, algorithms, or models that rely on machine learning or artificial intelligence to perform auscultation in a manner that significantly outperforms traditional approaches that rely on visual analysis by a healthcare professional.

Abstract: An adaptive active training system includes a motion module, a sensing module and a control module. The motion module includes a training unit and a motor connected to the training unit. The motor is configured to bring the training unit to move along a motion trajectory. The sensing module is configured to sense a physiological signal of a user when the user uses the training unit. The control module is connected to the motion module and the sensing module. The control module is configured to calculate a position of the training unit on the motion trajectory, obtain a threshold value corresponding to the position based on a motion model, and determine whether a magnitude of the physiological signal is greater than the threshold value.

Abstract: A method for scene perception using video captioning based on a spatio-temporal graph model is described. The method includes decomposing the spatio-temporal graph model of a scene in input video into a spatial graph and a temporal graph. The method also includes modeling a two branch framework having an object branch and a scene branch according to the spatial graph and the temporal graph to learn object interactions between the object branch and the scene branch. The method further includes transferring the learned object interactions from the object branch to the scene branch as privileged information. The method also includes captioning the scene by aligning language logits from the object branch and the scene branch according to the learned object interactions.

Abstract: An adaptive active training system includes a motion module, a sensing module and a control module. The motion module includes a training unit and a motor connected to the training unit. The motor is configured to bring the training unit to move along a motion trajectory. The sensing module is configured to sense a physiological signal of a user when the user uses the training unit. The control module is connected to the motion module and the sensing module. The control module is configured to calculate a position of the training unit on the motion trajectory, obtain a threshold value corresponding to the position based on a motion model, and determine whether a magnitude of the physiological signal is greater than the threshold value.

Abstract: The present application provides route planning methods and apparatuses for an unmanned aerial vehicle. An exemplary route planning method may include determining a quadrant angle in accordance with a starting point and an ending point. The route planning method may also include determining a flight shifting distance in accordance with the quadrant angle, a first interval between waypoints, and a second interval between routes. The route planning method may further include generating a plurality of waypoints and a plurality of routes in accordance with the quadrant angle, the flight shifting distance, the first interval between waypoints, and the second interval between routes. In addition, the route planning method include planning a flight path in accordance with the plurality of waypoints and the plurality of routes.

Abstract: Apparatuses, systems, and techniques to identify a goal of a demonstration. In at least one embodiment, video data of a demonstration is analyzed to identify a goal. Object trajectories identified in the video data are analyzed with respect to a task predicate satisfied by a respective object trajectory, and with respect to motion predicate. Analysis of the trajectory with respect to the motion predicate is used to assess intentionality of a trajectory with respect to the goal.

Abstract: An assembling method is provided for assembling a camera module with a substrate and a lens module includes the following steps. In a step (A), a sensing chip is installed on a surface of the substrate. In a step (B), a fixture is assembled with the substrate, wherein plural inner walls of the fixture are arranged around at least a lateral edge of the substrate. In a step (C), a glue is coated on a periphery of the surface of the substrate, wherein a movable range of the glue is limited by the plural inner walls of the fixture. In a step (D), a lens holder of the lens module is installed on the surface of the substrate, so that the lens holder and the substrate are combined together through the glue. In a step (E), the fixture is separated from the substrate.

Abstract: A dynamic motion detection system including a plurality of active independent emitting elements, a signal capturing apparatus, and a computing apparatus is provided. The active independent emitting elements are suitable for being respectively affixed to different parts of a person to-be-tested and actively emitting positioning signals having preset wavelengths. The signal capturing apparatus is configured to capture the positioning signal of each of the active independent emitting elements and calculate a plurality of emitting coordinates according to the positioning signals. The computing apparatus is configured to control operation of the signal capturing apparatus and receive the emitting coordinates from the signal capturing apparatus. The computing apparatus defines a corresponding relationship between the emitting coordinates and the active independent emitting elements by comparing a geometry relationship among the emitting coordinates.

Abstract: A monitoring system includes an image capturing device arranged to generate an image data of a scene; and a coordinate generating device arranged to calculate a coordinate of an object in the scene according to the image data.

Abstract: Embodiments of the present application provide information processing and model training methods, apparatuses, electronic devices and storage mediums.

Abstract: An assembling method is provided for assembling a camera module with a substrate and a lens module includes the following steps. In a step (A), a sensing chip is installed on a surface of the substrate. In a step (B), a fixture is assembled with the substrate, wherein plural inner walls of the fixture are arranged around at least a lateral edge of the substrate. In a step (C), a glue is coated on a periphery of the surface of the substrate, wherein a movable range of the glue is limited by the plural inner walls of the fixture. In a step (D), a lens holder of the lens module is installed on the surface of the substrate, so that the lens holder and the substrate are combined together through the glue. In a step (E), the fixture is separated from the substrate.

Abstract: A monitoring system includes an image capturing device arranged to generate an image data of a scene; and a coordinate generating device arranged to calculate a coordinate of an object in the scene according to the image data.

Abstract: The present application provides route planning methods and apparatuses for an unmanned aerial vehicle. An exemplary route planning method may include determining a quadrant angle in accordance with a starting point and an ending point. The route planning method may also include determining a flight shifting distance in accordance with the quadrant angle, a first interval between waypoints, and a second interval between routes. The route planning method may further include generating a plurality of waypoints and a plurality of routes in accordance with the quadrant angle, the flight shifting distance, the first interval between waypoints, and the second interval between routes. In addition, the route planning method include planning a flight path in accordance with the plurality of waypoints and the plurality of routes.

Abstract: A monitoring apparatus includes: a first operational device arranged to perform a first predetermined function and accordingly transmit a first instruction signal; a second operational device arranged to receive a second instruction signal and accordingly perform a second predetermined function; a first monitoring device coupled to the first operational device for generating a first detecting event according to an operation of the first operational device; and a second monitoring device coupled to the second operational device for generating a second detecting event according to the operation of the second operational device. The first monitoring device is wirelessly coupled to the second monitoring device, and the first detecting event and the second detecting event are used to determine if the first operational device and the second operational device perform the first predetermined function and the second predetermined function respectively.

Abstract: A dynamic motion detection system including a plurality of active independent emitting elements, a signal capturing apparatus, and a computing apparatus is provided. The active independent emitting elements are suitable for being respectively affixed to different parts of a person to-be-tested and actively emitting positioning signals having preset wavelengths. The signal capturing apparatus is configured to capture the positioning signal of each of the active independent emitting elements and calculate a plurality of emitting coordinates according to the positioning signals. The computing apparatus is configured to control operation of the signal capturing apparatus and receive the emitting coordinates from the signal capturing apparatus. The computing apparatus defines a corresponding relationship between the emitting coordinates and the active independent emitting elements by comparing a geometry relationship among the emitting coordinates.

Abstract: A modular light control device and a dimming control system are disclosed. The modular light control device comprises a power conversion module, an output module and a control module. The power conversion module has a conversion circuit and a first recognition unit. The conversion circuit receives a power source and converts the power source into a power signal. The output module has an output circuit and a second recognition unit. When the control module is electrically connected with the power conversion module and the output module, the control module distinguishes configurations of the power conversion module and the output module through the first recognition unit and the second recognition unit, respectively, and outputs a control signal according to the configuration of the output module, and the output circuit outputs a driving signal to drive a lamp according to the power signal and the control signal.

Abstract: The present invention is a method for preparing a biocontrol formulation by the following steps. A Streptomyces bacterium is selected simultaneously with superior sporulation ability, chitinase activity and antagonistic effectiveness against plant fungal pathogens by culturing with fungal pathogens, wherein the selected Streptomyces bacterium is cultured on a chitin-limited plate to assess the ability of sporulation and chitinase activity, and the Streptomyces bacterium with superior sporulation ability is the strain with producing powder-like and maturity form spore chains. Spores from the selected bacterium is collected. The obtained spores are taken as a seed inoculum. Then the seed inoculum is cultured to directly yield a broth medium containing at least 109 viable spores/ml which the spores contained in the broth medium are well suspended. The biocontrol formulation is obtained. The present invention further provided a biocontrol formulation containing a high concentration of Streptomyces spp. spores.

Abstract: Methods of treating a disease or disorder characterized by cells with increased or aberrant expression of cytokeratin-8 (CK8) are disclosed. The methods typically include administering to a subject in need thereof a pharmaceutical composition including an effective amount of a CK8 inhibitor. Exemplary diseases include thyroid cancers, particularly thyroid cancers characterized by poorly differentiated or undifferentiated cells. In the most preferred embodiments the cancer is an anaplastic thyroid cancer or a poorly differentiated papillary thyroid cancer. Disclosed inhibitors include, functional nucleic acids, inhibitory anti-CK8 antibodies, inhibitory peptides, and small molecules.