CPC G06T 5/002 (2013.01) [G06T 3/4007 (2013.01); G06T 3/4046 (2013.01); G06V 10/454 (2022.01); G06V 10/60 (2022.01); G06V 10/82 (2022.01); G06V 10/993 (2022.01); G06T 2207/20084 (2013.01)] | 17 Claims |
1. A method, comprising:
receiving a digital image;
generating, by at least one processor, a resulting digital image by processing the digital image with an encoder-decoder neural network comprising a plurality of convolutional layers classified into a downsampling stage and an upsampling stage, and a multi-scale context aggregating block configured to aggregate multi-scale context information of the digital image and employed between the downsampling stage and the upsampling stage; and
outputting, by the at least one processor, the resulting digital image to an output device,
wherein the generating the resulting digital image comprises:
concatenating the convolutional layers of the downsampling stage and the convolutional layers of the upsampling stage having a same resolution with the convolutional layers of the downsampling stage;
extracting, by a global pooling layer of the multi-scale context aggregating block, global context information of the digital image; and
extracting, by a plurality of dilation layers with various dilation rates of the multi-scale context aggregation block, context information of the digital image at different scales.
|