| CPC G06V 30/414 (2022.01) [G06F 18/21 (2023.01); G06N 3/08 (2013.01); G06V 10/95 (2022.01); G06V 30/18 (2022.01); G06V 30/274 (2022.01)] | 20 Claims |
|
10. A system comprising:
a semantic module implemented at least partially in hardware of a computing device to:
receive input data describing a digital image depicting a form; and
generate estimate data describing estimated associations of elements included in the form by processing the input data using a hierarchical convolutional neural network trained on training data to receive an image depicting an input form as an input and generate indications of associations of elements included in the input form as an output;
an embedding module implemented at least partially in the hardware of the computing device to:
extract an image patch from the digital image, the image patch depicts a pair of elements of the elements included in the form; and
encode an indication of whether the pair of elements have an association of the estimated associations; and
a multimodal module implemented at least partially in the hardware of the computing device to generate an indication that the pair of elements have a particular association based at least partially on the encoded indication, bounding boxes of the pair of elements, and text depicted in the image patch.
|