CPC G06N 3/045 (2023.01) [G06T 3/40 (2013.01); G06T 5/30 (2013.01); G06T 5/77 (2024.01); G06T 7/11 (2017.01); G06T 7/143 (2017.01); G06T 7/187 (2017.01); G06T 7/62 (2017.01); G06T 2207/20084 (2013.01)] | 20 Claims |
1. A computer-implemented method, comprising:
receiving an input image;
generating, by a first trained neural network model, a global probability map representation of the input image indicating, for each pixel of the input image, a probability value representing a likelihood of the pixel including a representation of wires;
identifying, based on the global probability map, regions of the input image indicated as including the representation of wires;
for each region of the input image from the identified regions of the input image,
concatenating the region of the input image and information from the global probability map to create a concatenated input, and
generating, by a second trained neural network model, a local probability map representation of the region of the input image based on the concatenated input, the local probability map indicating pixels of the region of the input image including representations of wires; and
aggregating local probability maps for each region of the input image, the aggregated local probability maps indicating pixels of the input image including the representations of wires.
|