Abstract: A method of compositional feature representation learning for video understanding is described. The method includes individually processing a sequence of video frames received as an input of a feature map network to generate a plurality of feature maps. The method also includes binding the plurality of feature maps to a fixed set of slot variables using an attention model according to a motion segmentation signal. The method further includes combining slot states corresponding to the fixed set of slot variables into a combined feature map. The method also includes decoding the combined feature map to form a reconstructed sequence of video frames, in which objects discovered in the reconstructed sequence of video frames are identified.

Abstract: An interventional operation isolation transfer cabin includes a cabin body, and a lead wire access device and a wire in/out device which two are mounted on the cabin body, the cabin body is provided with a negative pressure generating device at one end and an air inlet at the other end, the negative pressure generating device is provided with an exhaust filter device at its air outlet, and an intake filter device is provided at the air inlet. Several operating openings are provided on side walls of the cabin body, each of the operating openings is in seal connection with an operating glove, and a glove opening of at least one operating glove faces an interior of the cabin body. Equipment wires required by an interventional operation are preset through the lead wire access device, and wires added midway under a sealing condition is guaranteed through the wire in/out device.