Abstract: A personal care system having a treatment device for applying a treatment to the skin or hair of a user is provided. The treatment device has a camera for taking an image of the skin. The system may include or be in communication with an application programming interface (API) that can process the image. The system or an associated API can recommend the use of treatment regimens or topical products according to information determined from the digital image of the skin.

Abstract: At least one example embodiment relates to a computer-implemented method for determining an imaging parameter value for the control of a medical technology device during a capture of a first image dataset. In this context, the first image dataset is provided to be transferred from the medical technology device to a remotely arranged device. The method comprises at least one of receiving or determining a transfer parameter value. In this context, the transfer parameter value comprises an information item concerning which image information is relevant for the first image dataset which is to be transferred. The method further comprises determining the imaging parameter value based on the transfer parameter value. The method further comprises providing the imaging parameter value.

Abstract: An example method includes detecting, by context analysis circuitry, occurrence of a triggering condition. The example method also includes scheduling, by context analysis circuitry and based on the occurrence of the triggering condition, retraining of a model. The example method also includes generating, by data grafting circuitry and in response to scheduling the retraining of the model, a context-relevant training data set based on a target context vector. The example method also includes retraining, by model training circuitry, the model using the context-relevant training data set to mitigate deterioration of performance of the model.

Abstract: An example method includes detecting, by context analysis circuitry, occurrence of a triggering condition. The example method also includes scheduling, by context analysis circuitry and based on the occurrence of the triggering condition, retraining of a model. The example method also includes generating, by data grafting circuitry and in response to scheduling the retraining of the model, a context-relevant training data set based on a target context vector. The example method also includes retraining, by model training circuitry, the model using the context-relevant training data set to mitigate deterioration of performance of the model.

Abstract: Described herein are systems, methods, and instrumentalities associated with processing complex-valued MRI data using a machine learning (ML) model. The ML model may be learned based on synthetically generated MRI training data and by applying one or more meta-learning techniques. The MRI training data may be generated by adding phase information to real-valued MRI data and/or by converting single-coil MRI data into multi-coil MRI data based on coil sensitivity maps. The meta-learning process may include using portions of the training data to conduct a first round of learning to determine updated model parameters and using remaining portions of the training data to test the updated model parameters. Losses associated with the testing may then be determined and used to refine the model parameters. The ML model learned using these techniques may be adopted for a variety of tasks including, for example, MRI image reconstruction and/or de-noising.

Abstract: A surgical medical imaging system includes a second image processing device that acquires a first surgical image signal being a surgical image signal subjected to first image processing, from a medical imaging device including an imaging device and a first image processing device performing the first image processing on the surgical image signal captured by the imaging device, and performs second image processing on the first surgical image signal; and a third image processing device that acquires the first surgical image signal and a second surgical image signal being a surgical image signal subjected to the second image processing, performs third image processing on at least one of the first surgical image signal and the second surgical image signal, and generates a display image signal by conversion processing based on the first and the second surgical image signals, in which the first, second, and third image processing are different.

Abstract: In some examples, an apparatus, such as a mobile phone may include a camera, a processor, and computer readable medium. The camera capture one or more images of a person. The processor may use a machine learning model to predict the body volume of the person based on the captured images. The model may be trained based at a training data set comprising at least a plurality of user images. The apparatus may transmit the predicted body volume to a medication and medical treatment management system and receive from the same an adjusted medication or medical treatment plan. The apparatus may further execute the adjusted medication or medical treatment plan.

Abstract: A computer-implemented neural network system including a first machine learning system, in particular a first neural network, a second machine learning system, in particular a second neural network, and a third machine learning system, in particular a third neural network. The first machine learning system is designed to ascertain a higher-dimensional constructed image from a predefinable low-dimensional latent variable. The second machine learning system is designed to ascertain the latent variable again from the higher-dimensional constructed image, and the third machine learning system is designed to distinguish whether or not an image it receives is a real image.

Abstract: Embodiments include systems and methods that may be performed by a processor of a computing device. Embodiments may be applied for keypoint detection in an image. In embodiments, the processor of the computing device may apply to an image a first-stage neural network to define and output a plurality of regions, apply to each of the plurality of regions a respective second-stage neural network to output a plurality of keypoints in each of the plurality of regions, and apply to the plurality of keypoints a third-stage neural network to determine a correction for each of the plurality of keypoints to provide corrected keypoints.

Abstract: Systems and methods for automatically determining an image quality assessment of a rendered medical image are provided. A rendered medical image is received. One or more measures of interest are extracted from the rendered medical image. An image quality assessment of the rendered medical image is determined using a machine learning based image quality assessment network based on the one or more measures of interest. The image quality assessment of the rendered medical image is output.

Abstract: A mobile device can implement a neural network-based style transfer scheme to modify an image in a first style to a second style. The style transfer scheme can be configured to detect an object in the image, apply an effect to the image, and blend the image using color space adjustments and blending schemes to generate a realistic result image. The style transfer scheme can further be configured to efficiently execute on the constrained device by removing operational layers based on resources available on the mobile device.

Abstract: An image processing apparatus for determining a focused output image in a passive autofocus system is configured to retrieve a set of input images and compute a baseline estimate for at least one input image. The baseline estimate represents image structures in the input image. The image structures have a length scale larger than a predetermined image feature length scale. The image processing apparatus is further configured to compute a set of output images, wherein each output image of the set of output images is computed based on one of a different input image of the set of input images and the at least one baseline estimate for the different input image and the at least one baseline estimate for a respective different input image. The image processing apparatus is further configured to determine one output image of the set of output images as the focused output image.

Abstract: Systems and methods for determining corresponding locations of points of interest in a plurality of input medical images are provided. A plurality of input medical images comprising a first input medical image and one or more additional input medical images is received. The first input medical image identifies a location of a point of interest. A set of features is extracted from each of the plurality of input medical images. Features between each of the sets of features are related using a machine learning based relational network. A location of the point of interest in each of the one or more additional input medical images that corresponds to the location of the point of interest in the first input medical image is identified based on the related features. The location of the point of interest in each of the one or more additional input medical images is output.

Abstract: In some embodiments, a method can include receiving first images of produce. The method can further include executing a first machine learning model to generate second images of produce based on the first images of produce. The first images of produce can include (1) images of non-bagged produce or (2) images of bagged produce. The second images of produce can include the other of (1) images of non-bagged produce or (2) images of bagged produce. The method can further include training a second machine learning model based on the first images of produce and the second images of produce. The method can further include executing, after the training, the second machine learning model to classify as a bagged produce or a non-bagged produce an image not included in the first images and not included in the second images.

Abstract: An image processing apparatus includes a processor including hardware. The processor is configured to: acquire a first wavelength image and a second wavelength image, the first wavelength image being generated by irradiating a stage and by capturing light that has passed through the stage, the second wavelength image being generated by irradiating a specimen and by capturing light that has passed through the stage and the specimen, the specimen including a core tissue; calculate a spectral transmittance image; cause a display to display at least one of the spectral transmittance image and the second wavelength image as a display image; extract an area of the spectral transmittance image having spectral transmittance similar to spectral transmittance of a reference area in the display image as a core tissue area of the core tissue; and calculate an amount of the core tissue area.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a synthesized signal. A computer-implemented system obtains generator input data including an input signal having one or more first characteristics, processes the generator input data to generate output data including a synthesized signal having one or more second characteristics using a generator neural network, and outputs the synthesized signal to a device. The generator neural network is trained, based on a plurality of training examples, with a discriminator neural network. The discriminator neural network is configured to process discriminator input data that combines a discriminator input signal having the one or more second characteristics with at least a portion of generator input data to generate a prediction of whether the discriminator input signal is a real signal or a synthesized signal.

Abstract: Medical image segmentation using interactive refinement, in which the trained deep models are then utilized for the processing of medical imaging are described. Operating a two-step deep learning training framework including receiving original input images at the deep learning training framework; generating an initial prediction image specifying image segmentation by base segmentation model; receiving user input guidance signals; routing each of (i) the original input images, (ii) the initial prediction image, and (iii) the user input guidance signals to an InterCNN; generating a refined prediction image specifying refined image segmentation by processing each of the (i) the original input images, (ii) the initial prediction image, and (iii) the user input guidance signals through the InterCNN to render the refined prediction image incorporating the user input guidance signals; and outputting a refined segmentation mask to the deep learning training framework as a guidance signal.

Abstract: An all-optical Diffractive Deep Neural Network (D2NN) architecture learns to implement various functions or tasks after deep learning-based design of the passive diffractive or reflective substrate layers that work collectively to perform the desired function or task. This architecture was successfully confirmed experimentally by creating 3D-printed D2NNs that learned to implement handwritten classifications and lens function at the terahertz spectrum. This all-optical deep learning framework can perform, at the speed of light, various complex functions and tasks that computer-based neural networks can implement, and will find applications in all-optical image analysis, feature detection and object classification, also enabling new camera designs and optical components that can learn to perform unique tasks using D2NNs. In alternative embodiments, the all-optical D2NN is used as a front-end in conjunction with a trained, digital neural network back-end.

Abstract: A smart dandruff analysis system and method are provided for analyzing a severity of a subject's dandruff, and the smart dandruff analysis system has operation module, first neural network module, second neural network module, and classification module. The operation module receives a scalp area image of the subject and transforms the scalp area image into a first feature map. The first neural network module, a Convolutional Neural Network model, electrically connects with the operation module for receiving and transforming the scalp area image into a second feature map. The second neural network module, a Transformer model, electrically connecting with the first neural network module for receiving and transforming the second feature map into a third feature map. The classification module electrically connects with the second neural network module for receiving the third feature map and outputting a rating, wherein the rating is to determine the severity of the subject's dandruff.

Abstract: Generating and providing a condition assessment for a habitat is described. A geographic area having one or more geographic sub-areas is received. For each geographic sub-area, image data is received, a set of features is generated using the image data, a habitat for the geographic sub-area is received, and one or more condition assessment criteria are identified based on the habitat. For each condition assessment criterion, a subset of features is determined, and a trained model is applied to the subset of features to obtain a response to the condition assessment criterion. For each geographic sub-area, a condition assessment for the habitat is generated based on one or more responses to the one or more condition assessment criteria, and the condition assessment is provided.