The present invention belongs to the technical field of 3D reconstruction in the field of computer vision, and provides a dataset generation method for self-supervised learning scene point cloud completion based on panoramas. Pairs of incomplete point cloud and target point cloud with RGB information and normal information can be generated by taking RGB panoramas, depth panoramas and normal panoramas in the same view as input for constructing a self-supervised learning dataset for training of the scene point cloud completion network. The key points of the present invention are occlusion prediction and equirectangular projection based on view conversion, and processing of the stripe problem and point-to-point occlusion problem during conversion. The method of the present invention includes simplification of the collection mode of the point cloud data in a real scene; occlusion prediction idea of view conversion; and design of view selection strategy.

The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately and efficiently learning parameters of a distilled neural network from parameters of a source neural network utilizing multiple augmentation strategies. For example, the disclosed systems can generate lightly augmented digital images and heavily augmented digital images. The disclosed systems can further learn parameters for a source neural network from the lightly augmented digital images. Moreover, the disclosed systems can learn parameters for a distilled neural network from the parameters learned for the source neural network. For example, the disclosed systems can compare classifications of heavily augmented digital images generated by the source neural network and the distilled neural network to transfer learned parameters from the source neural network to the distilled neural network via a knowledge distillation loss function.

Broadly speaking, the present techniques generally relate to a method for training a machine learning, ML, model to perform semantic image segmentation, and to a computer-implemented method and apparatus for performing semantic image segmentation using a trained machine learning, ML, model. The training method enables a semantic image segmentation ML model that is able to make predictions faster, without significant loss in accuracy. The training method also enables the ML model to be implemented on apparatus with different hardware specifications, i.e. different computational power and memory, for example.

An apparatus for recognizing an object in an image includes a preprocessing module configured to receive an image including an object and to output a preprocessed image by performing image enhancement processing on the received image to improve a recognition rate of the object included in the received image; and an object recognition module configured to recognize the object included in the image by inputting the preprocessed image to an input layer of an artificial neural network for object recognition.

Included are: a simulation data acquiring unit to acquire simulation sensor data and acquire simulation traveling data; a feature amount calculating unit to calculate a feature amount from the simulation sensor data; a hyperparameter evaluation unit to evaluate whether or not a hyperparameter is a determined hyperparameter by comparing the simulation traveling data with ideal traveling data; a hyperparameter determination control unit to reset the hyperparameter until the hyperparameter evaluation unit evaluates that the hyperparameter is the determined hyperparameter, and repeatedly operate a mobile object simulator; and a teacher data generating unit to generate teacher data in which the hyperparameter evaluated as the determined hyperparameter by the hyperparameter evaluation unit and the feature amount calculated by the feature amount calculating unit are paired.

Data is received that includes a feed of images of a plurality of objects passing in front of an inspection camera module forming part of an inspection system. Thereafter, a representation for each image is generated using a first machine learning model and based on the received data. Later, one or more second machine learning models can cluster the images using the corresponding representations into groups that each correspond to one of a plurality of different object attributes. Thereafter, access to the groups can be provided to a consuming application or process for analysis and the like. In some variations, the representations are analyzed by at least one third machine learning model prior to the clustering. In other variations, the representations are analyzed by at least one third machine learning model after the clustering. Related apparatus, systems, and methods are also described.

An image processing system includes an interface to which an annotation result on a learning image captured inside a living body is input and a processor including hardware. The processor acquires metadata including difficulty information indicating difficulty of the annotation of the learning image itself, determines reliability information indicating reliability of the annotation result based on the metadata, and outputs a dataset in which the learning image, the annotation result, and the reliability information are associated with each other, as data used in generating a trained model used in inference based on deep learning on an inference target image captured inside a living body.

A system and method for providing weakly-supervised online action segmentation that include receiving image data associated with multi-view videos of a procedure, wherein the procedure involves a plurality of atomic actions. The system and method also include analyzing the image data using weakly-supervised action segmentation to identify each of the plurality of atomic actions by using an ordered sequence of action labels. The system and method additionally include training a neural network with data pertaining to the plurality of atomic actions based on the weakly-supervised action segmentation. The system and method further include executing online action segmentation to label atomic actions that are occurring in real-time based on the plurality of atomic actions trained to the neural network.

Certain aspects of the present disclosure provide techniques and apparatus for processing video content using an artificial neural network. An example method generally includes receiving a video data stream including at least a first frame and a second frame. First features are extracted from the first frame using a teacher neural network. A difference between the first frame and the second frame is determined. Second features are extracted from at least the difference between the first frame and the second frame using a student neural network. A feature map for the second frame is generated based a summation of the first features and the second features. An inference is generated for at least the second frame of the video data stream based on the generated feature map for the second feature.

A false positive removal engine is provided. The false positive removal engine receives detected objects in one or more images. A machine learning classifier computer model, configured with first operational parameters to implement a first operating point, processes the received input to classify each detected object as being a true positive or a false positive to generate a first set of object classifications. If the first set is empty, the false positive removal engine outputs the first set as a filtered list of objects; otherwise the ML classifier computer model is configured with second operational parameters to implement a second operating point, different from the first operating point, which then processes the received input to classify each detected object and generate a second set of objects classified as true positive, which is output by the false positive removal engine as the filtered list of objects.