Abstract: The present disclosure provides a knob type rope adjuster, includes a base, a rope winding disc, a knob, and an unlocking button; the rope winding disc used for winding an adjustment rope is rotatably installed in the base; the knob is installed on the base in an unidirectional rotation mode; the knob is combined with the rope winding disc to drive the rope winding disc to rotate via the knob, thereby winding and locking the adjustment rope; the unlocking button is connected to the rope winding disc and passes through the knob, thereby switching combination or separation of the knob and the rope winding disc; the knob is separated from the rope winding disc, so that the rope winding disc is arranged to rotate in a direction opposite to the winding direction, thereby loosening the adjustment rope, the adjustment rope remains locked on the rope winding disc to avoid loosening.

Abstract: The present invention discloses a method for point cloud up-sampling based on deep learning, including: obtaining training data including a first number of sparse input points and a second number of dense input points; constructing a deep network model to be used for respectively performing replication and sampling operation based on curvature on initial eigenvectors extracted from the first number of sparse input points to obtain a second number of intermediate eigenvectors, performing splicing operation on each intermediate eigenvector, inputting the spliced intermediate eigenvectors into a multilayer perceptron, and determining sampling prediction points based on the sampling eigenvectors output by the multilayer perceptron; training the deep network model until an objective function determined by the sampling prediction points and the dense input points converges; and testing the deep network model to obtain point cloud data of an object under test after up-sampling.

Abstract: An adjustable neck support includes a neck holder front support, a left side plate, a right side plate, a lower jaw holder plate, and a height adjustment knob. The left side plate is movably connected at a left end of the neck holder front support, the right side plate is movably connected at a right end of the neck holder front support, the lower jaw holder plate is connected between the left side plate and the right side plate, and the height adjustment knob is mounted on the neck holder front support. The height adjustment knob includes an operating cover and a winding post connected to the operating cover, the winding post being mounted in the neck holder front support. The adjustable neck support provided by the application is smooth in height adjustment operation, easy to operate, and long in service life.

Abstract: An adjustable neck support includes a neck holder front support, a left side plate, a right side plate, a lower jaw holder plate, and a height adjustment knob. The left side plate is movably connected at a left end of the neck holder front support, the right side plate is movably connected at a right end of the neck holder front support, the lower jaw holder plate is connected between the left side plate and the right side plate, and the height adjustment knob is mounted on the neck holder front support. The height adjustment knob includes an operating cover and a winding post connected to the operating cover, the winding post being mounted in the neck holder front support. The adjustable neck support provided by the application is smooth in height adjustment operation, easy to operate, and long in service life.

Abstract: The present invention discloses a method for point cloud up-sampling based on deep learning, including: obtaining training data including a first number of sparse input points and a second number of dense input points; constructing a deep network model to be used for respectively performing replication and sampling operation based on curvature on initial eigenvectors extracted from the first number of sparse input points to obtain a second number of intermediate eigenvectors, performing splicing operation on each intermediate eigenvector, inputting the spliced intermediate eigenvectors into a multilayer perceptron, and determining sampling prediction points based on the sampling eigenvectors output by the multilayer perceptron; training the deep network model until an objective function determined by the sampling prediction points and the dense input points converges; and testing the deep network model to obtain point cloud data of an object under test after up-sampling.