Abstract: Embodiments of the disclosure provide a method and a device for processing a portrait image, the method includes: acquiring a to-be-processed portrait image; inputting the to-be-processed portrait image into an image processing model, and acquiring a head smear image output by the image processing model, where the image processing model is configured to smear a hair area of a portrait located above a preset boundary in the portrait image, and the image processing model is generated by training a sample data set of a sample portrait image and a sample head smear image corresponding to the sample portrait image; rendering the head smear image with a head effect material to obtain a portrait image added with an effect; and displaying the portrait image added with the effect.

Abstract: In implementations of object detection in images, object detectors are trained using heterogeneous training datasets. A first training dataset is used to train an image tagging network to determine an attention map of an input image for a target concept. A second training dataset is used to train a conditional detection network that accepts as conditional inputs the attention map and a word embedding of the target concept. Despite the conditional detection network being trained with a training dataset having a small number of seen classes (e.g., classes in a training dataset), it generalizes to novel, unseen classes by concept conditioning, since the target concept propagates through the conditional detection network via the conditional inputs, thus influencing classification and region proposal. Hence, classes of objects that can be detected are expanded, without the need to scale training databases to include additional classes.

Abstract: A video generation method includes: acquiring an original image corresponding to a target frame, and identifying a target object in the original image; according to a sliding zooming strategy, performing sliding zooming processing on an initial background image in the original image excluding the target object, so as to obtain a target background image, wherein the sliding zooming strategy is at least used for indicating a sliding direction and a zooming direction of the initial background image, and the sliding direction is opposite to the zooming direction; according to the position of the target object in the original image, superimposing an image of the target object onto the target background image to obtain a target image corresponding to the target frame; and generating a target video on the basis of the target image corresponding to the target frame.

Abstract: Provided are a transition type determination method, an electronic device and a storage medium. The method includes: acquiring a picture matching degree between a candidate transition type and a transition position of two adjacent video clips, and acquiring a music matching degree of the candidate transition type and background music of a video to which the two adjacent video clips belong; and determining a target transition type for the transition position according to the picture matching degree and the music matching degree, where the target transition type is used for a transition effect between the two adjacent video clips.

Abstract: Embodiments of the present invention provide an automated image tagging system that can predict a set of tags, along with relevance scores, that can be used for keyword-based image retrieval, image tag proposal, and image tag auto-completion based on user input. Initially, during training, a clustering technique is utilized to reduce cluster imbalance in the data that is input into a convolutional neural network (CNN) for training feature data. In embodiments, the clustering technique can also be utilized to compute data point similarity that can be utilized for tag propagation (to tag untagged images). During testing, a diversity based voting framework is utilized to overcome user tagging biases. In some embodiments, bigram re-weighting can down-weight a keyword that is likely to be part of a bigram based on a predicted tag set.

Abstract: The present disclosure describes techniques of detecting objects in a video. The techniques comprises extracting features from each frame of the video; generating a first attentive feature by applying a first attention model on at least some of features extracted from any particular frame among the plurality of frames, wherein the first attention model identifies correlations between a plurality of locations in the particular frame by computing relationships between any two locations among the plurality of locations; generating a second attentive feature by applying a second attention model on at least one pair of features at different levels selected from the features extracted from the particular frame, wherein the second attention model identifies a correlation between at least one pair of locations corresponding to the at least one pair of features; and generating a representation of an object included in the particular frame.

Abstract: Provided are a transition type determination method, an electronic device and a storage medium. The method includes: acquiring a picture matching degree of a candidate transition type with a transition position between two adjacent video clips, and acquiring a music matching degree of the candidate transition type; determining, based on the acquired image matching degree and the acquired music matching degree, a matching degree of the candidate transition type at the transition position; and determining, according to the matching degree, whether to determine the candidate transition type as a target transition type, where the target transition type is used for a transition effect between the two adjacent video clip.

Abstract: Embodiments of the present application provide an image display method and apparatus, a device, and a storage medium. The method comprises: displaying a preceding image in a first display period, the preceding image comprising a video image sequence or a single image; superimposing a foreground target area of a succeeding image on an upper layer of the preceding image in a second display period for display, the succeeding image comprising a video image sequence or a single image; and displaying the succeeding image in a third display period. According to the method, a good playback effect can be achieved in scenarios where time variations are desired.

Abstract: Embodiments of the present invention are directed to facilitating region of interest preservation. In accordance with some embodiments of the present invention, a region of interest preservation score using adaptive margins is determined. The region of interest preservation score indicates an extent to which at least one region of interest is preserved in a candidate image crop associated with an image. A region of interest positioning score is determined that indicates an extent to which a position of the at least one region of interest is preserved in the candidate image crop associated with the image. The region of interest preservation score and/or the preserving score are used to select a set of one or more candidate image crops as image crop suggestions.

Abstract: A video generation method includes: acquiring an original image corresponding to a target frame, and identifying a target object in the original image; according to a sliding zooming strategy, performing sliding zooming processing on an initial background image in the original image excluding the target object, so as to obtain a target background image, wherein the sliding zooming strategy is at least used for indicating a sliding direction and a zooming direction of the initial background image, and the sliding direction is opposite to the zooming direction; according to the position of the target object in the original image, superimposing an image of the target object onto the target background image to obtain a target image corresponding to the target frame; and generating a target video on the basis of the target image corresponding to the target frame.

Abstract: Provided are a transition type determination method, an electronic device and a storage medium. The method includes: acquiring a picture matching degree of a candidate transition type with a transition position between two adjacent video clips, and acquiring a music matching degree of the candidate transition type; determining, based on the acquired image matching degree and the acquired music matching degree, a matching degree of the candidate transition type at the transition position; and determining, according to the matching degree, whether to determine the candidate transition type as a target transition type, where the target transition type is used for a transition effect between the two adjacent video clip.

Abstract: The present disclosure relates to training and utilizing an image exposure transformation network to generate a long-exposure image from a single short-exposure image (e.g., still image). In various embodiments, the image exposure transformation network is trained using adversarial learning, long-exposure ground truth images, and a multi-term loss function. In some embodiments, the image exposure transformation network includes an optical flow prediction network and/or an appearance guided attention network. Trained embodiments of the image exposure transformation network generate realistic long-exposure images from single short-exposure images without additional information.

Abstract: The present disclosure describes techniques of detecting objects in a video. The techniques comprises extracting features from each frame of the video; generating a first attentive feature by applying a first attention model on at least some of features extracted from any particular frame among the plurality of frames, wherein the first attention model identifies correlations between a plurality of locations in the particular frame by computing relationships between any two locations among the plurality of locations; generating a second attentive feature by applying a second attention model on at least one pair of features at different levels selected from the features extracted from the particular frame, wherein the second attention model identifies a correlation between at least one pair of locations corresponding to the at least one pair of features; and generating a representation of an object included in the particular frame.

Abstract: Embodiments of the present application provide an image display method and apparatus, a device, and a storage medium. The method comprises: displaying a preceding image in a first display period, the preceding image comprising a video image sequence or a single image; superimposing a foreground target area of a succeeding image on an upper layer of the preceding image in a second display period for display, the succeeding image comprising a video image sequence or a single image; and displaying the succeeding image in a third display period. According to the method, a good playback effect can be achieved in scenarios where time variations are desired.

Abstract: Semantic segmentation techniques and systems are described that overcome the challenges of limited availability of training data to describe the potentially millions of tags that may be used to describe semantic classes in digital images. In one example, the techniques are configured to train neural networks to leverage different types of training datasets using sequential neural networks and use of vector representations to represent the different semantic classes.

Abstract: Systems and techniques for estimating illumination from a single image are provided. An example system may include a neural network. The neural network may include an encoder that is configured to encode an input image into an intermediate representation. The neural network may also include an intensity decoder that is configured to decode the intermediate representation into an output light intensity map. An example intensity decoder is generated by a multi-phase training process that includes a first phase to train a light mask decoder using a set of low dynamic range images and a second phase to adjust parameters of the light mask decoder using a set of high dynamic range image to generate the intensity decoder.

Abstract: Predicting patch displacement maps using a neural network is described. Initially, a digital image on which an image editing operation is to be performed is provided as input to a patch matcher having an offset prediction neural network. From this image and based on the image editing operation for which this network is trained, the offset prediction neural network generates an offset prediction formed as a displacement map, which has offset vectors that represent a displacement of pixels of the digital image to different locations for performing the image editing operation. Pixel values of the digital image are copied to the image pixels affected by the operation.

Abstract: Various embodiments describe frame selection based on training and using a neural network. In an example, the neural network is a convolutional neural network trained with training pairs. Each training pair includes two training frames from a frame collection. The loss function relies on the estimated quality difference between the two training frames. Further, the definition of the loss function varies based on the actual quality difference between these two frames. In a further example, the neural network is trained by incorporating facial heatmaps generated from the training frames and facial quality scores of faces detected in the training frames. In addition, the training involves using a feature mean that represents an average of the features of the training frames belonging to the same frame collection. Once the neural network is trained, a frame collection is input thereto and a frame is selected based on generated quality scores.

Abstract: In implementations of object detection in images, object detectors are trained using heterogeneous training datasets. A first training dataset is used to train an image tagging network to determine an attention map of an input image for a target concept. A second training dataset is used to train a conditional detection network that accepts as conditional inputs the attention map and a word embedding of the target concept. Despite the conditional detection network being trained with a training dataset having a small number of seen classes (e.g., classes in a training dataset), it generalizes to novel, unseen classes by concept conditioning, since the target concept propagates through the conditional detection network via the conditional inputs, thus influencing classification and region proposal. Hence, classes of objects that can be detected are expanded, without the need to scale training databases to include additional classes.

Abstract: Digital image completion by learning generation and patch matching jointly is described. Initially, a digital image having at least one hole is received. This holey digital image is provided as input to an image completer formed with a dual-stage framework that combines a coarse image neural network and an image refinement network. The coarse image neural network generates a coarse prediction of imagery for filling the holes of the holey digital image. The image refinement network receives the coarse prediction as input, refines the coarse prediction, and outputs a filled digital image having refined imagery that fills these holes. The image refinement network generates refined imagery using a patch matching technique, which includes leveraging information corresponding to patches of known pixels for filtering patches generated based on the coarse prediction. Based on this, the image completer outputs the filled digital image with the refined imagery.