公开(公告)号：WO2022098488A1

公开(公告)日：2022-05-12

申请号：PCT/US2021/055157

申请日：2021-10-15

Applicant: INNOPEAK TECHNOLOGY, INC.

Inventor： LI, Jiachen ,  LIU, Rongrong ,  LIN, Yuan

IPC: G06K9/00 ,  G06N20/00

Abstract: This application is directed to identifying text areas in an image. A computer system obtains the image including one or more text areas, and generates a sequence of feature maps from the image based on a downsampling rate. Each feature map has a first dimension and a second dimension, and the feature maps include a first feature map and a second feature map. Each of the first and second dimensions of the first feature map has a respective size that is reduced to that of a respective dimension of the second feature map by the downsampling rate. The second feature map is upsampled by an upsampling rate using a local context-aware upsampling network. The upsampled second feature map is aggregated with the first feature map to generate an aggregated first feature map. The one or more text areas are identified in the image based on the aggregated first feature map.

公开(公告)号：WO2022099325A1

公开(公告)日：2022-05-12

申请号：PCT/US2022/011790

申请日：2022-01-10

Applicant: INNOPEAK TECHNOLOGY, INC.

Inventor： LI, Jiachen ,  ZHANG, Kaiyu ,  LIU, Rongrong ,  LIN, Yuan

IPC: G06K9/62 ,  G06T7/00 ,  G06K9/00

Abstract: A system may include a backbone network configured to generate feature maps from an image, a transformer network coupled to the backbone network, and a scene text detection subsystem, the scene text detection subsystem comprising a processor, and a non-transitory computer readable medium having encoded thereon a set of instructions executable by the processor to generate a plurality of image tokens from one or more feature maps of an input image, and generate, via the transformer encoder, a set of token queries, wherein the set of token queries quantify an attention of a respective textual feature of a respective image token of the sequence of image tokens relative to all other respective textual features of all other image tokens, and generate, via the transformer decoder, a set of predicted text boxes of the input image.