Abstract: The disclosure provides a method for enhancing audio-visual association by adopting self-supervised curriculum learning. With the help of contrastive learning, the method can train the visual and audio model without human annotation and extracts meaningful visual and audio representations for a variety of downstream tasks in the context of a teacher-student network paradigm. Specifically, a two-stage self-supervised curriculum learning scheme is proposed to contrast the visual and audio pairs and overcome the difficulty of transferring between visual and audio information in the teacher-student framework. Moreover, the knowledge shared between audio and visual modality serves as a supervisory signal for contrastive learning. In summary, with the large-scale unlabeled data, the method can obtain a visual and an audio convolution encoder. The encoders are helpful for downstream tasks and cover the training shortage causing by limited data.

Abstract: A power-split hydro-mechanical hybrid transmission system with an automatic adjustment function includes an input member, a hydraulic transmission mechanism, a split mechanism, a convergence mechanism, an output member, a clutch assembly, and a brake assembly. The clutch assembly connects the input member to an input end of the split mechanism, connects an output end of the split mechanism to an input end of the hydraulic transmission mechanism and an input end of the convergence mechanism, and connects an output end of the hydraulic transmission mechanism to the output member. An output end of the convergence mechanism is connected to the output member. The clutch assembly and the brake assembly provide a continuous transmission ratio between the input member and the output member. The power-split hydro-mechanical hybrid transmission system enables multi-mode continuously variable transmission and has energy reuse and emergency support functions.

Abstract: A multi-source domain adaptation model by aligning partial features includes a general feature extraction module, a feature selection module for partial feature extraction with the dedicated loss function, three partial feature alignment losses, and two classifiers for adversarial training, where the three partial feature alignment losses include an intra-class partial feature alignment loss, an inter-domain partial feature alignment loss, and an inter-class partial feature alignment loss. With the partial features extracted by the general feature extraction module and the feature selection module following three different partial feature alignment losses, the model is capable of clustering the samples from the identical categories and isolating samples from distinct classes.

Abstract: A power-split hydro-mechanical hybrid transmission system with an automatic adjustment function includes an input member, a hydraulic transmission mechanism, a split mechanism, a convergence mechanism, an output member, a clutch assembly, and a brake assembly. The clutch assembly connects the input member to an input end of the split mechanism, connects an output end of the split mechanism to an input end of the hydraulic transmission mechanism and an input end of the convergence mechanism, and connects an output end of the hydraulic transmission mechanism to the output member. An output end of the convergence mechanism is connected to the output member. The clutch assembly and the brake assembly provide a continuous transmission ratio between the input member and the output member. The power-split hydro-mechanical hybrid transmission system enables multi-mode continuously variable transmission and has energy reuse and emergency support functions.

Abstract: A hydro-mechanical hybrid transmission device with two hydraulic transmission mechanisms includes an input member, a second hydraulic transmission mechanism, a rear planetary gear mechanism, an output member, a first hydraulic transmission mechanism, a front planetary gear mechanism, a clutch assembly, and a brake assembly. The clutch assembly connects the input member to the front planetary gear mechanism, the second hydraulic transmission mechanism, and the first hydraulic transmission mechanism, connects an output end of the first hydraulic transmission mechanism to the front planetary gear mechanism, connects the rear planetary gear mechanism to an output end of the second hydraulic transmission mechanism and the front planetary gear mechanism, and connects the rear planetary gear mechanism to the output member. The clutch assembly and the brake assembly provide a continuous transmission ratio between the input member and the output member.

Abstract: Provided are a time-of-flight depth camera and a noise-reduction distance measurement method. The depth camera comprises: a light source for emitting a pulse beam to an object to be measured; an image sensor comprising at least one pixel, wherein each of the at least one pixel comprises taps, and the taps are used for acquiring a charge signal generated by a reflected pulse beam reflected by the object to be measured and/or a charge signal of background light; and a processing circuit, configured to: control the taps to alternately acquire charge signals in frame periods of a macro period, wherein different modulation and demodulation frequencies are used in two adjacent macro periods; and receive data of charge signals acquired in the two adjacent macro periods to calculate a time of flight of the pulse beam and/or a distance from the depth camera to the object to be measured.

Abstract: A multi-mode continuously variable transmission with both speed coupling and torque coupling includes an engine-power input assembly, a hydraulic transmission assembly, a motor transmission assembly, a planetary gear assembly, an output member, a clutch assembly, and a brake assembly, wherein an output end of the planetary gear assembly is connected to the output member, the clutch assembly connects the engine-power input assembly, the hydraulic transmission assembly, and the motor transmission assembly to an input end of the planetary gear assembly, and the clutch assembly connects the engine-power input assembly to the hydraulic transmission assembly; and the clutch assembly and the brake assembly provide a continuously changing transmission ratio between the engine-power input assembly or/and the motor transmission assembly and the output member.

Abstract: A structured light projection module, a depth camera, and a method for manufacturing the structured light projection module are provided. The module comprises: a light source, comprising a plurality of sub-light sources that are arranged in a two-dimensional array and configured to emit two-dimensional patterned beams corresponding to the two-dimensional array, and the two-dimensional patterned beams comprising two-dimensional patterns; a lens, receiving and converging the two-dimensional patterned beams; and a diffractive optical element, receiving the two-dimensional patterned beams converged and emitted from the lens, and projecting speckle patterned beams corresponding to speckle patterns. The speckle patterns comprise a plurality of image patterns corresponding to the two-dimensional patterns, and the image patterns are rotated by an angle such that edges of the image patterns are not in parallel with a baseline between the structured light projection module and a capture module.

Abstract: A hydro-mechanical hybrid transmission device with two hydraulic transmission mechanisms includes an input member, a second hydraulic transmission mechanism, a rear planetary gear mechanism, an output member, a first hydraulic transmission mechanism, a front planetary gear mechanism, a clutch assembly, and a brake assembly. The clutch assembly connects the input member to the front planetary gear mechanism, the second hydraulic transmission mechanism, and the first hydraulic transmission mechanism, connects an output end of the first hydraulic transmission mechanism to the front planetary gear mechanism, connects the rear planetary gear mechanism to an output end of the second hydraulic transmission mechanism and the front planetary gear mechanism, and connects the rear planetary gear mechanism to the output member. The clutch assembly and the brake assembly provide a continuous transmission ratio between the input member and the output member.

Abstract: A multi-mode continuously variable transmission with both speed coupling and torque coupling includes an engine-power input assembly, a hydraulic transmission assembly, a motor transmission assembly, a planetary gear assembly, an output member, a clutch assembly, and a brake assembly, wherein an output end of the planetary gear assembly is connected to the output member, the clutch assembly connects the engine-power input assembly, the hydraulic transmission assembly, and the motor transmission assembly to an input end of the planetary gear assembly, and the clutch assembly connects the engine-power input assembly to the hydraulic transmission assembly; and the clutch assembly and the brake assembly provide a continuously changing transmission ratio between the engine-power input assembly or/and the motor transmission assembly and the output member.

Abstract: A diffractive optical element is configured to receive a first patterned light beams, and diffracts the first patterned light beams to project a second patterned light beam outwardly. The second patterned light beam is formed of combined arrangement of a plurality of first patterned light beams, and the combined arrangement includes periodic arrangement in a first direction and a second direction. The first direction and the second direction are not perpendicular to each other. By designing the performance of the diffractive optical element, the outwardly projected structured light pattern is periodically arranged in two directions, and the two directions are not perpendicular to each other, such that spot patterns are highly unrelated in multiple directions, improving accuracy.

Abstract: A structured light projection module, a depth camera, and a method for manufacturing the structured light projection module are provided. The module comprises: a light source, comprising a plurality of sub-light sources that are arranged in a two-dimensional array and configured to emit two-dimensional patterned beams corresponding to the two-dimensional array, and the two-dimensional patterned beams comprising two-dimensional patterns; a lens, receiving and converging the two-dimensional patterned beams; and a diffractive optical element, receiving the two-dimensional patterned beams converged and emitted from the lens, and projecting speckle patterned beams corresponding to speckle patterns. The speckle patterns comprise a plurality of image patterns corresponding to the two-dimensional patterns. Relationships between adjacent image patterns of the plurality of image patterns comprise at least two of an overlapping relationship, an adjoining relationship, and a spacing relationship.

Abstract: A hydro-mechanical transmission device with a dual-clutch transmission includes an input mechanism, a pump-controlled motor mechanism, an odd-numbered gear transmission mechanism, an even-numbered gear transmission mechanism, an output mechanism, and a jackshaft. The input mechanism is connected to the odd-numbered gear transmission mechanism and the even-numbered gear transmission mechanism that are connected in parallel, and is connected to an input end of the pump-controlled motor mechanism. An output end of the pump-controlled motor mechanism is connected, through the jackshaft, to the odd-numbered gear transmission mechanism and the even-numbered gear transmission mechanism that are connected in parallel, and is connected to the output mechanism. Switching among hydraulic transmission, hydro-mechanical transmission, and mechanical transmission modes between the input mechanism and the output mechanism is implemented by controlling combination and engagement/disengagement of clutches and a brake.

Abstract: A method may include identifying a time window of a procedure. The method may also include obtaining operational information of the time window. The operational information may include a limit of pulse repetition frequency (PRF) acceleration and a plurality of preliminary radio frequency (RF) PRFs. The method may also include determining a plurality of updated RF PRFs by updating the plurality of preliminary RF PRFs. A rate of variation between any two adjacent updated RF PRFs may be less than or equal to the limit of PRF acceleration. The method may also include causing an RF source to generate electromagnetic waves at the plurality of updated RF PRFs in the time window.

Abstract: A continuously variable transmission system for hybrid power multi-mode switching, includes an input component, an output component, a clutch assembly, a brake, a hydraulic transmission assembly and a planetary gear assembly, wherein the input component is connected with the hydraulic transmission assembly, the output component is connected with the planetary gear assembly, the clutch assembly connects the input component and the hydraulic transmission assembly to the planetary gear assembly respectively, and the brake and the clutch assembly provide a transmission ratio for continuous forwarding or backwarding continuously between the input component and the output component. The hydro-mechanical transmission is switched to mechanical transmission by increasing the displacement ratio of the hydraulic transmission assembly linearly or non-linearly.

Abstract: A touch panel having dummy pattern is provided, including a first metal nanowire layer and a second metal nanowire layer. The first metal nanowire layer includes first electrode wires, first axial wires connected to the first electrode wires, and first dummy patterns. The second metal nanowire layer includes a plurality of second electrode wires and a plurality of second axial wires connected to the second electrode wires. The first dummy patterns are electrically insulated and deposited outside the first electrode wires and the first axial wires, and each of the first dummy patterns comprises a plurality of first etching areas extending along the first and second directions. The first dummy patterns do not expose the first etching areas along the first direction at a vertical projection area of a part where each of the second axial wires alone exists, so that the electrode pattern is difficult to observe.

Abstract: A depth calculation processor, a data processing method, and a 3D image device are disclosed herein. The depth calculation processor includes: at least two input ports configured to receive a first image data, wherein the first image data comprises at least a structured light image acquired under projection of a structured light; a data processing engine configured to perform a calculation process on the first image data to output a second image data, wherein the second image data comprises at least a depth map, wherein the data processing engine comprises at least a depth processing engine configured to perform a depth calculation process on the structured light image to obtain the depth map; and at least an output port coupled to the data processing engine and configured to output the second image data to a host device.

Abstract: A depth calculation processor, a data processing method, and a 3D image device are disclosed herein. The depth calculation processor includes: two input ports to receive a first image data, wherein the first image data comprises structured light image acquired under projection of structured light; an input switch connected to the input ports and to convey all or some of the first image data from the input ports; a data processing engine connected to the input switch and to process the first image data that is output through the input switch and to output a second image data, wherein the second image data comprises a depth map, wherein the data processing engine comprises a depth processing engine to process the structured light image to obtain the depth map; and one output port connected to the data processing engine and to output the second image data to a main device.

Abstract: The present disclosure provides a structured light projection module comprising: a Vertical-Cavity Surface-Emitting Laser (VCSEL) array light source that includes a semiconductor substrate and at least two VCSEL sub-arrays arranged thereon. The at least two VCSEL sub-arrays include VCSEL light sources. The structured light projection module further includes a diffractive optical element (DOE) that includes at least two DOE sub-units. The DOE sub-units are respectively corresponding to the VCSEL sub-arrays and configured to project multiple copies of light beams emitted by the corresponding at least two VCSEL sub-arrays.