Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that modifies regions of a digital image via localized style transfer. For example, in some embodiments, the disclosed systems receive a natural language text input for modifying a digital image and determine, from the natural language text input, a first style for modifying a first region of the digital image and a second style for modifying a second region of the digital image. Additionally, the disclosed systems modify, using a multi-region style transfer neural network, the digital image by incorporating the first style within the first region and incorporating the second style within the second region. Further, the disclosed systems provide the modified digital image for display on a graphical user interface of a client device.

Abstract: An image generation system accessing an input image displayed on a user interface. The image generation system receives, via the user interface, a target style text defining a target style for a stylized image to be generated based on the input image and a request to generate the stylized image. The image generation system generates the stylized image. Generating the stylized image includes applying a text guided image generation model to the input image and the target style text, wherein the text guided image generation model minimizes a loss between a first relationship between the generated stylized image and a set of style templates and a second relationship between the target style text and the set of style templates. The image generation system displays, via the user interface responsive to receiving the request, the generated stylized image.

Abstract: Systems and methods for data augmentation are provided. One aspect of the systems and methods include receiving an image that is misclassified by a classification network; computing an augmentation image based on the image using an augmentation network; and generating an augmented image by combining the image and the augmentation image, wherein the augmented image is correctly classified by the classification network.

Abstract: An image generation system accessing an input image displayed on a user interface. The image generation system receives, via the user interface, a target style text defining a target style for a stylized image to be generated based on the input image and a request to generate the stylized image. The image generation system generates the stylized image. Generating the stylized image includes applying a text guided image generation model to the input image and the target style text, wherein the text guided image generation model minimizes a loss between a first relationship between the generated stylized image and a set of style templates and a second relationship between the target style text and the set of style templates. The image generation system displays, via the user interface responsive to receiving the request, the generated stylized image.

Abstract: A corrective noise system receives an electronic version of a fillable form generated by a segmentation network and receives a correction to a segmentation error in the electronic version of the fillable form. The corrective noise system is trained to generate noise that represents the correction and superimpose the noise on the fillable form. The corrective noise system is further trained to identify regions in a corpus of forms that are semantically similar to a region that was subject to the correction. The generated noise is propagated to the semantically similar regions in the corpus of forms and the noisy corpus of forms is provided as input to the segmentation network. The noise causes the segmentation network to accurately identify fillable regions in the corpus of forms and output a segmented version of the corpus of forms having improved fidelity without retraining or otherwise modifying the segmentation network.

Abstract: A corrective noise system receives an electronic version of a fillable form generated by a segmentation network and receives a correction to a segmentation error in the electronic version of the fillable form. The corrective noise system is trained to generate noise that represents the correction and superimpose the noise on the fillable form. The corrective noise system is further trained to identify regions in a corpus of forms that are semantically similar to a region that was subject to the correction. The generated noise is propagated to the semantically similar regions in the corpus of forms and the noisy corpus of forms is provided as input to the segmentation network. The noise causes the segmentation network to accurately identify fillable regions in the corpus of forms and output a segmented version of the corpus of forms having improved fidelity without retraining or otherwise modifying the segmentation network.

Abstract: Embodiments are disclosed for one-shot document snippet search. A method of one-shot document snippet search may include obtaining a query snippet and a target document. A multi-modal snippet detection model combines first multi-modal features from the query snippet and second multi-modal features from the target document to create a feature volume. The multi-modal snippet detection model identifies one or more matching snippets from the target document based on the feature volume.

Abstract: A corrective noise system receives an electronic version of a fillable form generated by a segmentation network and receives a correction to a segmentation error in the electronic version of the fillable form. The corrective noise system is trained to generate noise that represents the correction and superimpose the noise on the fillable form. The corrective noise system is further trained to identify regions in a corpus of forms that are semantically similar to a region that was subject to the correction. The generated noise is propagated to the semantically similar regions in the corpus of forms and the noisy corpus of forms is provided as input to the segmentation network. The noise causes the segmentation network to accurately identify fillable regions in the corpus of forms and output a segmented version of the corpus of forms having improved fidelity without retraining or otherwise modifying the segmentation network.

Abstract: Form structure similarity detection techniques are described. A content processing system, for instance, receives a query snippet that depicts a query form structure. The content processing system generates a query layout string that includes semantic indicators to represent the query form structure and generates candidate layout strings that represent form structures from a target document. The content processing system calculates similarity scores between the query layout string and the candidate layout strings. Based on the similarity scores, the content processing system generates a target snippet for display that depicts a form structure that is structurally similar to the query form structure. The content processing system is further operable to generate a training dataset that includes image pairs of snippets depicting form structures that are structurally similar.

Abstract: Form structure similarity detection techniques are described. A content processing system, for instance, receives a query snippet that depicts a query form structure. The content processing system generates a query layout string that includes semantic indicators to represent the query form structure and generates candidate layout strings that represent form structures from a target document. The content processing system calculates similarity scores between the query layout string and the candidate layout strings. Based on the similarity scores, the content processing system generates a target snippet for display that depicts a form structure that is structurally similar to the query form structure. The content processing system is further operable to generate a training dataset that includes image pairs of snippets depicting form structures that are structurally similar.

Abstract: Systems and methods for data augmentation are provided. One aspect of the systems and methods include receiving an image that is misclassified by a classification network; computing an augmentation image based on the image using an augmentation network; and generating an augmented image by combining the image and the augmentation image, wherein the augmented image is correctly classified by the classification network.