Computer vision systems and methods for localizing image forgery are provided. The system generates a constrained convolution via a plurality of learned rich filters. The system trains a convolutional neural network with the constrained convolution and a plurality of images of a dataset to learn a low level representation of each image among the plurality of images. The low level representation is indicative of a statistical signature of at least one source camera model of each image. The system can determine a splicing manipulation localization by the trained convolutional neural network.

Provided is an object detection device including: a detection unit configured to detect objects for every detection period to output detection information containing a reliability for each of the detected objects; a determination unit configured to: increment a detection count for each of the objects; calculate, for each of the objects, a sum of latest N reliabilities in the detection period; and determine, as a normally recognized object, an object for which the sum is equal to or larger than a first threshold value, which is set in advance depending on the detection count; and a control unit configured to output, as normally detected object information, detection information on the normally recognized object.

According to one embodiment, a learning method for causing a second statistical model to learn using a first statistical model is provided. The method includes obtaining a first learning image, cutting out each local area of the obtained first learning image, and obtaining a first prediction value output from the first statistical model by inputting each local area to the first statistical model and obtaining a second prediction value output from the second statistical model by inputting the entire area of the first learning image to the second statistical model, and causing the second statistical model to learn based on a difference between the first prediction value and the second prediction value.

To provide a robust information processing system against attacks by Adversarial Example. A neural network model 608, a latent space database 609 for storing position information in a latent space in which first output vectors, which are output vectors of a predetermined hidden layer included in the neural network model, are embedded concerning input data used for learning of the neural network model, and an inference control unit 606 for making an inference using the neural network model and the latent space database are provided. The inference control unit infers the input data based on the positional relationship between the second output vector, which is an output vector of the predetermined hidden layer concerning input data to be inferred, and the first output vectors in said latent space.

A system for image classification is disclosed that includes a central system configured to provide high reliability image data processing and recognition and a plurality of endpoint systems, each configured to provide image data processing and recognition with a lower reliability than the central system and to generate probability data. A decision switch disposed at each of the plurality of endpoint systems is configured to receive the probability data and to determine whether to deny access, grant access or generate a referral message to the central system, wherein the referral message includes at least a set of image data generated at the endpoint system.

A mechanism is provided in a data processing system comprising a processor and a memory, the memory comprising instructions executed by the processor to specifically configure the processor to implement a statistical model tool for providing insight into decision making. The statistical model tool applies the statistical model to an input image to generate an original classification probability. An image modification component executing within the statistical model tool iterative modifies each portion of the input image to generate a modified image. The statistical model tool applies the statistical model to the modified image to generate a new classification probability for each portion of the input image. A compare component executing in the statistical model tool compares each new classification probability to the original classification probability to generate a respective probability distance. A distance map generator executing within the statistical model tool generates a distance map data structure based on the probability distances. The distance map data structure represents an impact each portion of the input image has on determining classification probability by the statistical model.

Example methods and systems for generating an aggregated artificial intelligence (AI) engine for radiotherapy treatment planning are provided. One example method may include obtaining multiple AI engines associated with respective multiple treatment planners; generating multiple sets of output data using the multiple AI engines associated with the respective multiple treatment planners: comparing the multiple AI engines associated with the respective multiple treatment planners based on the multiple sets of output data; and based on the comparison, aggregating at least some of the multiple AI engines to generate the aggregated AI engine for performing the particular treatment planning step. The multiple AI engines may be trained to perform a particular treatment planning step, and each of the multiple AI engines is trained to emulate one of the multiple treatment planners performing the particular treatment planning step.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network system used to control an agent interacting with an environment to perform a specified task. One of the methods includes causing the agent to perform a task episode in which the agent attempts to perform the specified task; for each of one or more particular time steps in the sequence: generating a modified reward for the particular time step from (i) the actual reward at the time step and (ii) value predictions at one or more time steps that are more than a threshold number of time steps after the particular time step in the sequence; and training, through reinforcement learning, the neural network system using at least the modified rewards for the particular time steps.

Bias in Machine Learning (ML) is when an ML algorithm tends to incompletely learn relevant and important patterns from a dataset, or learns the patterns from data incorrectly. Such inaccuracy can cause the algorithm to miss important relationships between patterns and features in data, resulting in inaccurate algorithm predictions. Systems and methods for detecting potential ML bias in input image datasets are described herein. After a target image is received, a subset of images related to the target image is extracted. The target image and subset of images are analyzed under an imbalance assessment and data bias assessment to determine the presence of any potential data bias in a ML training pipeline. If any data bias is determined, one or more messages summarizing the assessments and including explanations to enable more accurate predictions in image assessments are sent to the user.

A system for image classification is disclosed that includes a central system configured to provide high reliability image data processing and recognition and a plurality of endpoint systems, each configured to provide image data processing and recognition with a lower reliability than the central system and to generate probability data. A decision switch disposed at each of the plurality of endpoint systems is configured to receive the probability data and to determine whether to deny access, grant access or generate a referral message to the central system, wherein the referral message includes at least a set of image data generated at the endpoint system.