An image processing apparatus comprises: an image input unit configured to input an image; a detection unit configured to detect an object from the image; an accepting unit configured to accept an input of a locus to the image, a selection unit configured to select, based on a locus region decided by the locus, at least two objects included in a plurality of objects detected by the detection unit; and an integration unit configured to generate an integration region that integrates at least two regions in the image corresponding to the at least two objects selected by the selection unit and set the integration region as a region of interest in the image.

A method for training a model based on knowledge distillation includes: inputting feature vectors obtained based on trained sample images into a first coding layer and a second coding layer, in which the first coding layer belongs to a first model, and the second coding layer belongs to a second model; obtaining first feature vectors by aggregating output results of the first coding layer; determining second feature vectors based on outputs of the second coding layer; and updating the first feature vectors by performing a distillation on the first feature vectors and the second feature vectors.

Disclosed herein is an image analysis system including a machine learning model configured to receive input of an image and to output an image feature quantity and a map source for image analysis, a map generation section configured to generate, on the basis of a vector corresponding to an object and the output map source, an attention map indicative of a region associated with the object in the image, and a token generation section configured to generate a token indicative of a feature associated with an event of the object on the basis of the generated attention map and the image feature quantity.

Disclosed herein is an image analysis system including a machine learning model configured to receive input of an image and to output an image feature quantity and a map source for image analysis, a map generation section configured to generate, on the basis of a vector corresponding to an object and the output map source, an attention map indicative of a region associated with the object in the image, and a token generation section configured to generate a token indicative of a feature associated with an event of the object on the basis of the generated attention map and the image feature quantity.

Techniques are described that facilitate training an artificial intelligence (AI) model in a manner that ensures equitable model performance across different sub-groups. According to an embodiment, a system is provided that comprises a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory. The computer executable components comprise a training component that trains a machine learning (ML) model on training data to perform an inferencing task using an equitable loss function that drives equitable performance of the ML model across different sub-groups represented the training data, resulting in trained version of the ML model that provides a defined equitable performance level across the different sub-groups. The equitable loss function is “sub-group aware” and penalizes variation in model performance across the sub-groups during model training and validation.

The present invention provides a flexible, light-weighted and efficient multistream fusion encoder which can be easily integrated into multistream convolutional neural networks to perform segmentation and classification tasks on MRI images registered with different modalities. The encoder allows fusion of extracted feature maps in multiple streams on a layer-by-layer basis and generates the output of each stream by adding the corresponding convolutional output with an adaptively weighted fusion map computed from outputs of all streams. Adaptive weighting of fusion maps at each layer allows flexibility in highlighting different image modalities according to their relative influence on the segmentation/classification performance. The fusion encoder can also play an important role in the segmentation-classification workflow in biopsy and focal therapy planning.

The present invention provides a flexible, light-weighted and efficient multistream fusion encoder which can be easily integrated into multistream convolutional neural networks to perform segmentation and classification tasks on MRI images registered with different modalities. The encoder allows fusion of extracted feature maps in multiple streams on a layer-by-layer basis and generates the output of each stream by adding the corresponding convolutional output with an adaptively weighted fusion map computed from outputs of all streams. Adaptive weighting of fusion maps at each layer allows flexibility in highlighting different image modalities according to their relative influence on the segmentation/classification performance. The fusion encoder can also play an important role in the segmentation-classification workflow in biopsy and focal therapy planning.

There is provided a computer implemented method of training a preanalytical factor machine learning model, comprising: creating a preanalytical training dataset of a plurality of records, wherein a preanalytical record comprises: an image of a slide of pathological tissue of a subject processed with at least one preanalytical factor, and a ground truth label indicating the at least one preanalytical factor, and training the preanalytical machine learning model on the preanalytical training dataset for generating an outcome of at least one target preanalytical factor used to process tissue depicted in a target image in response to the input of the target image.

There is provided a computer implemented method of training a preanalytical factor machine learning model, comprising: creating a preanalytical training dataset of a plurality of records, wherein a preanalytical record comprises: an image of a slide of pathological tissue of a subject processed with at least one preanalytical factor, and a ground truth label indicating the at least one preanalytical factor, and training the preanalytical machine learning model on the preanalytical training dataset for generating an outcome of at least one target preanalytical factor used to process tissue depicted in a target image in response to the input of the target image.

A system for detecting clouds and cloud shadows is described. In one approach, clouds and cloud shadows within a remote sensing image are detected through a three step process. In the first stage a high-precision low-recall classifier is used to identify cloud seed pixels within the image. In the second stage, a low-precision high-recall classifier is used to identify potential cloud pixels within the image. Additionally, in the second stage, the cloud seed pixels are grown into the potential cloud pixels to identify clusters of pixels which have a high likelihood of representing clouds. In the third stage, a geometric technique is used to determine pixels which likely represent shadows cast by the clouds identified in the second stage. The clouds identified in the second stage and the shadows identified in the third stage are then exported as a cloud mask and shadow mask of the remote sensing image.