Abstract: Embodiments are disclosed for generating an instant mask from polarized input images.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for training and utilizing scale-diverse segmentation neural networks to analyze digital images at different scales and identify different target objects portrayed in the digital images. For example, in one or more embodiments, the disclosed systems analyze a digital image and corresponding user indicators (e.g., foreground indicators, background indicators, edge indicators, boundary region indicators, and/or voice indicators) at different scales utilizing a scale-diverse segmentation neural network. In particular, the disclosed systems can utilize the scale-diverse segmentation neural network to generate a plurality of semantically meaningful object segmentation outputs. Furthermore, the disclosed systems can provide the plurality of object segmentation outputs for display and selection to improve the efficiency and accuracy of identifying target objects and modifying the digital image.

Abstract: This disclosure describes one or more implementations of an alpha matting system that utilizes a deep learning model to generate alpha mattes for digital images utilizing an alpha-range classifier function. More specifically, in various implementations, the alpha matting system builds and utilizes an object mask neural network having a decoder that includes an alpha-range classifier to determine classification probabilities for pixels of a digital image with respect to multiple alpha-range classifications. In addition, the alpha matting system can utilize a refinement model to generate the alpha matte from the pixel classification probabilities with respect to the multiple alpha-range classifications.

Abstract: Embodiments are disclosed for performing universal segmentation to mask objects across multiple frames of a video. The method may include determining an image segmentation mask which masks an object of a frame of a video sequence using the frame and an image segmentation module of a segmentation system. The method further includes determining a mask propagation mask which masks the object of the frame of the video sequence using the frame, a representation of a previous frame of the video sequence, and a mask propagation module of the segmentation system. The method further includes determining a frame mask which masks the object of the frame of the video sequence based on a comparison of the image segmentation mask and the mask propagation mask. The method further includes displaying the frame mask of the video sequence.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing a plurality of neural networks in a multi-branch pipeline to generate image masks for digital images. Specifically, the disclosed system can classify a digital image as a portrait or a non-portrait image. Based on classifying a portrait image, the disclosed system can utilize separate neural networks to generate a first mask portion for a portion of the digital image including a defined boundary region and a second mask portion for a portion of the digital image including a blended boundary region. The disclosed system can generate the mask portion for the blended boundary region by utilizing a trimap generation neural network to automatically generate a trimap segmentation including the blended boundary region. The disclosed system can then merge the first mask portion and the second mask portion to generate an image mask for the digital image.

Abstract: A method includes receiving a frame depicting an object. The frame is one frame of a plurality of frames of a video sequence. The method further includes encoding a plurality of tokens of the frame. Each token is a representation of a grid of pixels of the frame. The method further includes selecting a subset of tokens for decoding based on a likelihood of a token satisfying a confidence threshold. The token satisfies the confidence threshold based on a confidence score of the token including a past object in a past frame. The method further includes decoding the subset of tokens using a decoder.

Abstract: A method includes receiving an object mask of an object in an image. The method further includes generating a mask of a sub-object in the image using a machine learning model configured to receive the mask of the object. A first branch of the machine learning model predicts whether a pixel of the image belongs to a sub-object.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing a transformer-based encoder-decoder neural network architecture for generating alpha mattes for digital images. Specifically, the disclosed system utilizes a transformer encoder to generate patch-based encodings from a digital image and a trimap segmentation by generating patch encodings for image patches and comparing the patch encodings to areas of the digital image. Additionally, the disclosed system generates modified patch-based encodings utilizing a plurality of neural network layers. The disclosed system also generates an alpha matte for the digital image from the patch-based encodings utilizing a decoder that includes a plurality of upsampling layers connected to a plurality of neural network layers via skip connections.

Abstract: Embodiments are disclosed for a machine learning-based chroma keying process. The method may include receiving an input including an image depicting a chroma key scene and a color value corresponding to a background color of the image. The method may further include generating a preprocessed image by concatenating the image and the color value. The method may further include providing the preprocessed image to a trained neural network. The method may further include generating, using the trained neural network, an alpha matte representation of the image based on the preprocessed image.

Abstract: Systems and methods perform image matte generation using image bursts. In accordance with some aspects, an image burst comprising a set of images is received. Features of a reference image from the set of images is aligned with features of other images from the set of images. A matte for the reference image is generated using the aligned features.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for efficiently automating the preparation of accurate alpha matte animations and modified digital videos utilizing polarized light. For example, the disclosed systems obtain a plurality of polarized digital videos portraying an animation of a foreground subject backlit by a polarized light source. In some embodiments, the disclosed systems generate a plurality of corrected polarized digital videos by adjusting intensity values of the plurality of polarized digital videos based on intensity differences across the plurality of polarized digital videos. The disclosed systems generate an alpha matte animation comprising a plurality of alpha mattes from the plurality of corrected polarized digital videos or from the plurality of polarized digital videos.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing a transformer-based encoder-decoder neural network architecture for generating alpha mattes for digital images. Specifically, the disclosed system utilizes a transformer encoder to generate patch-based encodings from a digital image and a trimap segmentation by generating patch encodings for image patches and comparing the patch encodings to areas of the digital image. Additionally, the disclosed system generates modified patch-based encodings utilizing a plurality of neural network layers. The disclosed system also generates an alpha matte for the digital image from the patch-based encodings utilizing a decoder that includes a plurality of upsampling layers connected to a plurality of neural network layers via skip connections.

Abstract: Embodiments are disclosed for predicting a user's intent to generate a masked region of an image based on a user gesture. The method may include receiving a user interaction corresponding to a region in an image. The user interaction is used to select the object in the image. The method further includes generating a feature map using a first machine learning model and the user interaction, where the feature map includes at least one feature corresponding to a mode type. The mode type is an intended operation of a user. Lastly, the method includes generating a mask of the object in the image using a second machine learning model, wherein the mask is generated responsive to the mode type.

Abstract: Systems, methods, and non-transitory computer-readable media embed a trained neural network within a digital image. For instance, in one or more embodiments, the systems identify out-of-gamut pixel values of a digital image in a first gamut, where the digital image is converted to the first gamut from a second gamut. Furthermore, the systems determine target pixel values of a target version of the digital image in the first gamut that correspond to the out-of-gamut pixel values. The systems train a neural network to predict the target pixel values in the first gamut based on the out-of-gamut pixel values. The systems embed the neural network within the digital image in the second gamut to allow for extraction of the embedded neural network from the digital image to restore the digital image to a larger gamut digital image.

Abstract: The present disclosure relates to a multi-model object segmentation system that provides a multi-model object segmentation framework for automatically segmenting objects in digital images. In one or more implementations, the multi-model object segmentation system utilizes different types of object segmentation models to determine a comprehensive set of object masks for a digital image. In various implementations, the multi-model object segmentation system further improves and refines object masks in the set of object masks utilizing specialized object segmentation models, which results in more improved accuracy and precision with respect to object selection within the digital image. Further, in some implementations, the multi-model object segmentation system generates object masks for portions of a digital image otherwise not captured by various object segmentation models.

Abstract: The present invention relates to a method for producing a three-dimensional (3D) printed article with a photocurable silicone composition involving a silicone containing as end-group specific (meth)acrylate groups.

Abstract: The present disclosure relates to a class-agnostic object segmentation system that automatically detects, segments, and selects objects within digital images irrespective of object semantic classifications. For example, the object segmentation system utilizes a class-agnostic object segmentation neural network to segment each pixel in a digital image into an object mask. Further, in response to detecting a selection request of a target object, the object segmentation system utilizes a corresponding object mask to automatically select the target object within the digital image. In some implementations, the object segmentation system utilizes a class-agnostic object segmentation neural network to detect and automatically select a partial object in the digital image in response to a target object selection request.

Abstract: A curable silicone adhesive having improved elongation-at-break and adhesive properties to various substrates, in particular synthetic textiles used in the manufacture of air bags, to be used, for example, as a joint sealer. These silicone compositions provide excellent adhesive properties such as peel strength and cohesive failure when used to seal joints/seams between two pieces of textile fabric. Airbag fabrics using such novel addition curable adhesive silicone compositions are also provided.

Abstract: An image processing system auto white balances an image using an object in the image and a reference color distribution. Given an input image, a target object in the input image is identified. A reference color distribution for the object type of the target object from the input image is accessed. One or more image processing settings are determined that, when applied to the input image, minimize a difference in values between pixels of the target object and the reference color distribution. A white balanced image is generated by applying the one or more image processing settings to the input image, and the white balanced image is provided for presentation.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for performing interactive digital image editing operations utilizing machine learning models and a feature backpropagation refinement layer. For example, the disclosed systems perform interactive digital image editing operations by incorporating a feature backpropagation refinement layer within a non-interactive machine learning model that utilizes a consistency loss to adjust the feature backpropagation refinement layer according to one or more user interactions. In some embodiments, the disclosed systems utilize a feature backpropagation refinement layer that includes a bias sublayer for localizing changes to a digital image and a convolutional sublayer for channel-wise scale and feature combinations across channels. In some cases, the disclosed systems utilize a consistency loss that facilitates localized modifications to a digital image based on distances of various pixels or features from a user interaction.