Abstract: A video stream processing method and apparatus are provided. The method includes obtaining an image set of a target person from a multi-channel video stream, the multi-channel video stream being obtained for a same scene by a plurality of cameras, and an image in the image set includes a front face image of the target person; determining a virtual viewpoint in a target-person view mode based on the image in the image set; and projecting, based on a depth map of a target image and a pose of a real viewpoint corresponding to the target image, the target image onto an imaging plane corresponding to the virtual viewpoint to obtain a video stream in the target-person view mode, the target image intersecting with a vision field of the target person in the multi-channel video stream.

Abstract: An image processing model training method includes inputting image data in a training dataset into an image processing model to perform processing, to obtain a processing result corresponding to the image data, where parameters of n1 neurons are scaled up, and parameters of n2 neurons are scaled down in the image processing model, calculating an error between an annotation result of the image data in the training dataset and the processing result, and adjusting parameters of the image processing model based on the error between the annotation result and the processing result, where n1 and n2 are positive integers.

Abstract: A video stream processing method and apparatus are provided. The method includes obtaining an image set of a target person from a multi-channel video stream, the multi-channel video stream being obtained for a same scene by a plurality of cameras, and an image in the image set includes a front face image of the target person; determining a virtual viewpoint in a target-person view mode based on the image in the image set; and projecting, based on a depth map of a target image and a pose of a real viewpoint corresponding to the target image, the target image onto an imaging plane corresponding to the virtual viewpoint to obtain a video stream in the target-person view mode, the target image intersecting with a vision field of the target person in the multi-channel video stream.

Abstract: The present disclosure provides a method for semiconductor manufacturing in accordance with some embodiments. The method includes providing a substrate and a patterning layer over the substrate and forming a plurality of openings in the patterning layer. The substrate includes a plurality of features to receive a treatment process. The openings partially overlap with the features from a top view while a portion of the features remains covered by the patterning layer. Each of the openings is free of concave corners. The method further includes performing an opening expanding process to enlarge each of the openings and performing a treatment process to the features through the openings. After the opening expanding process, the openings fully overlap with the features from the top view.

Abstract: A panoramic camera is configured to automatically determine parameters of a table upon which the camera is situated as well as positional information of the camera relative to the table. In an initialization stage, table edges are detected to create an edge map. A Hough transformation-like symmetry voting operation is performed to clean up the edge map and to determine camera offset, camera orientation and camera tilt. The table is then fit to a table model to determine table parameters. In an operational stage, table edges are detected to create an edge map and the table model is fit to the edge map. The output can then be used for further panoramic image processing such as head size normalization, zooming, compensation for camera movement, etc.

Abstract: A panoramic camera is configured to automatically determine parameters of a table upon which the camera is situated as well as positional information of the camera relative to the table. In an initialization stage, table edges are detected to create an edge map. A Hough transformation-like symmetry voting operation is performed to clean up the edge map and to determine camera offset, camera orientation and camera tilt. The table is then fit to a table model to determine table parameters. In an operational stage, table edges are detected to create an edge map and the table model is fit to the edge map. The output can then be used for further panoramic image processing such as head size normalization, zooming, compensation for camera movement, etc.