公开(公告)号：US11387131B2

公开(公告)日：2022-07-12

申请号：US16810945

申请日：2020-03-06

Applicant: Kioxia Corporation

Inventor： Miki Toshima ,  Osamu Yamane

IPC: H01L25/065 ,  H01L23/544 ,  H01L21/68 ,  H01L25/00 ,  H01L25/18 ,  H01L21/66 ,  H01L23/00 ,  G06V10/22

Abstract: An alignment apparatus according to one embodiment, includes: a first and a second stage; a first and a second detector; a first and a second moving mechanism; and a controller. The first and second stages are configured to respectively hold a first and a second semiconductor substrate on which a first and a second alignment mark are respectively disposed. The first and second moving mechanisms are configured to respectively move the first and second stages relatively to each other. The controller is configured to perform the following (a), (b). (a) The controller control the detectors and the moving mechanisms to cause the first detector to detect the second alignment mark and to cause the second detector to detect the first alignment mark. (b) The controller calculate a position deviation between the substrates in accordance with results of the detections.

公开(公告)号：US11715194B2

公开(公告)日：2023-08-01

申请号：US17191063

申请日：2021-03-03

Applicant: Kioxia Corporation

Inventor： Youyang Ng ,  Bo Wang ,  Takuji Ohashi ,  Osamu Yamane ,  Takeshi Fujiwara

IPC: G06T7/00 ,  G06F18/214

CPC classification number: G06T7/001 ,  G06F18/2155 ,  G06T2207/20081 ,  G06T2207/30148

Abstract: An information processing apparatus has an acquisitor configured to acquire an entire area image obtained by capturing an entire area of a processing surface of a wafer including at least one defect, a training image selector configured to select, as a training image, a partial image including at least one defect from the entire area image, a model constructor configured to construct a calculation model of generating a label image obtained by extracting and binarizing the defect included in the partial image, and a learner configured to update a parameter of the calculation model based on a difference between the label image generated by inputting the training image to the calculation model and a reference label image obtained by extracting and binarizing the defect of the training image.