An electronic device is provided. The electronic device includes at least one processor and a memory. The memory may store at least one instruction that, when executed by the at least one processor, enables the electronic device to obtain a first image frame and a second image frame. The memory may store at least one instruction that, when executed by the at least one processor, enables the electronic device to identify a first interpolation-applied object and a first interpolation non-applied object among objects included in the first image frame and identify a second interpolation-applied object and a second interpolation non-applied object among objects included in the second image frame. The memory may store at least one instruction that, when executed by the at least one processor, enables the electronic device to provide an interpolation image frame including a result of interpolation on the first interpolation-applied object and the second interpolation-applied object and the first interpolation non-applied object or the second interpolation non-applied object.

A system of automatic optimization of image quality of an image sensor includes an image learning data generation unit generating an image tuning knowledge database, which includes pairs of a plurality of sets of values of a plurality of parameters and a plurality of sets of image quality evaluation scores for a plurality of image quality evaluation items for evaluating a quality of each of a plurality of images generated by the image sensor, using an image tuning database sampling module, an image signal processor modeling unit generating a machine learning model, for each image, for automatically optimizing the quality of each image, and an image sensor image quality optimization unit automatically controlling values of some of the plurality of parameters based on a user's image quality selection and the machine learning model. The image quality evaluation scores are produced by a distributed camera simulation system including servers.

A method include obtaining at least one first PET image of a subject acquired by a PET scanner and at least one first MR image of the subject acquired by an MR scanner. The method may also include obtaining a target neural network model. The target neural network model may provide a mapping relationship between PET images, MR images, and corresponding attenuation correction data, and output attenuation correction data associated with a specific PET image of the PET images. The method may further include generating first attenuation correction data corresponding to the subject using the target neural network model based on the at least one first PET image and the at least one first MR image of the subject, and determining a target PET image of the subject based on the first attenuation correction data corresponding to the subject.

In various embodiments, a computer-implemented method of training a neural network for relighting an image is described. A first training set that includes source images and a target illumination embedding is generated, the source images having respective illuminated subjects. A second training set that includes augmented images and the target illumination embedding is generated, where the augmented images corresponding to the source images. A first autoencoder is trained using the first training set to generate a first output set that includes estimated source illumination embeddings and first reconstructed images that correspond to the source images, the reconstructed images having respective subjects that are i) from the corresponding source image, and ii) illuminated based on the target illumination embedding. A second autoencoder is trained using the second training set to generate a second output set that includes estimated augmented illumination embeddings and second reconstructed images that correspond to the augmented images.

An apriori guidance network for multitask medical image synthesis is provided. The apriori guidance network includes a generator and a discriminator, wherein the generator includes an apriori guidance module configured to convert an input feature map into a target modal image pointing to a target domain according to an apriori feature, and the apriori feature is a deep feature of the target modal image. The generator is configured to generate a corresponding target domain image by taking the apriori feature of the target modal image and source modal image data as an input. The discriminator is configured to discriminate an authenticity of the target domain image outputted by the generator.

A computer-implemented method of reducing temporal motion artifacts in temporal intracardiac sensor data, includes: inputting (S120) temporal intracardiac sensor data (110), into a neural network (130) trained to predict, from the temporal intracardiac sensor data (110), temporal motion data (140, 150) representing the temporal motion artifacts (120); and compensating (S130) for the temporal motion artifacts (120) in the received 5 temporal intracardiac sensor data (110) based on the predicted temporal motion data (140, 150).

An image processing method and an image processing device are provided. The image processing method includes performing at least one of a scratch repairing step, a dead point repairing step, a denoising step and a color cast correcting step on a to-be-processed video frame. According to the embodiments of the present disclosure, it is able to repair a scratch and a dead point, remove a noise and/or correct color cast for a video frame, thereby to improve a display effect of the video frame.

A method for driving an electronic device is described. The method may include receiving light passing through a display panel, generating an image signal based on the light, compensating the image signal with a compensation algorithm to generate a compensated image signal, wherein the compensation algorithm is trained with training data including a first comparison image and a second comparison image, and displaying, on the display panel, a compensated image based on the compensated image signal, wherein the first comparison image is a target restoration image and the second comparison image is a composite image.

Disclosed is a multi-scale fusion defogging method based on a stacked hourglass network, including inputting a foggy image into a preset image defogging network; and outputting a fogless image after the foggy image is processed by the image defogging network. The image defogging network includes a 7?7 convolutional layer, a stacked hourglass module, a feature fusion, a multi-scale jump connection module, a 1?1 convolutional layer, a 3?3 convolutional layer, a hierarchical attention distillation module, the 3?3 convolutional layer and the 1?1 convolutional layer connected sequentially.

A system and method for determine a parameter for medical data processing are provided. The method may include obtaining sample data, the sample data may comprise at least one of projection data or a scanning parameter. The method may also include obtaining a first neural network model. The method may further include determining the parameter based on the sample data and the first neural network model. The parameter may comprise at least one of a correction coefficient or a noise reduction parameter.