A method for determining hard example sensor data inputs for training a task neural network is described. The task neural network is configured to receive a sensor data input and to generate a respective output for the sensor data input to perform a machine learning task. The method includes: receiving one or more sensor data inputs depicting a same scene of an environment, wherein the one or more sensor data inputs are taken during a predetermined time period; generating a plurality of predictions about a characteristic of an object of the scene; determining a level of inconsistency between the plurality of predictions; determining that the level of inconsistency exceeds a threshold level; and in response to the determining that the level of inconsistency exceeds a threshold level, determining that the one or more sensor data inputs comprise a hard example sensor data input.

Apparatuses and methods are provided for training a feature extraction model determining a loss function for use in unsupervised image segmentation. A method includes determining a clustering loss from an image; determining a weakly supervised contrastive loss of the image using cluster pseudo labels based on the clustering loss; and determining the loss function based on the clustering loss and the weakly supervised contrastive loss.

Neural network performance is improved in terms of training speed and/or accuracy by encoding (mapping) inputs to the neural network into a higher dimensional space via a hash function. The input comprises coordinates used to identify a point within a d-dimensional space (e.g., 3D space). The point is quantized and a set of vertex coordinates corresponding to the point are input to a hash function. For example, for d=3, space may be partitioned into axis-aligned voxels of identical size and vertex coordinates of a voxel containing the point are input to the hash function to produce a set of encoded coordinates. The set of encoded coordinates is used to lookup D-dimensional feature vectors in a table of size T that have been learned. The learned feature vectors are filtered (e.g., linearly interpolated, etc.) based on the coordinates of the point to compute a feature vector corresponding to the point.

Systems and methods disclosed herein use a first machine learning architecture and a second machine learning architecture where the first machine learning architecture executes on a first processor and receives a first image representing a mouth of a user, determines user feedback for outputting to the user based on a first machine learning model, and outputs the user feedback for capturing a second image representing the mouth of the user. The second machine learning architecture executes on a second processor and receives the first image and the second image, and generates a 3D model of at least a portion of a dental arch of the user based on the first image and the second image where the 3D model is generated based on a second machine learning model of the second machine learning architecture.

Systems and methods disclosed herein use a first machine learning architecture and a second machine learning architecture where the first machine learning architecture executes on a first processor and receives a first image representing a mouth of a user, determines user feedback for outputting to the user based on a first machine learning model, and outputs the user feedback for capturing a second image representing the mouth of the user. The second machine learning architecture executes on a second processor and receives the first image and the second image, and generates a 3D model of at least a portion of a dental arch of the user based on the first image and the second image where the 3D model is generated based on a second machine learning model of the second machine learning architecture.

A method of controlling an image acquisition device for tracking a target object includes: detecting an event in which tracking of a first object, which is a tracking target object, fails in a first image acquired by the image acquisition device; determining, in the first image, a reference object which is used as a reference for controlling the image acquisition device; controlling the image acquisition device such that at least one of an image capturing range and an image capturing direction of the image acquisition device is adjusted based on at least one of a size and a location of the reference object in the first image; and recognizing the first object in a second image acquired by the image acquisition device in a state in which at least one of the image capturing range and the image capturing direction is adjusted.

A method of controlling an image acquisition device for tracking a target object includes: detecting an event in which tracking of a first object, which is a tracking target object, fails in a first image acquired by the image acquisition device; determining, in the first image, a reference object which is used as a reference for controlling the image acquisition device; controlling the image acquisition device such that at least one of an image capturing range and an image capturing direction of the image acquisition device is adjusted based on at least one of a size and a location of the reference object in the first image; and recognizing the first object in a second image acquired by the image acquisition device in a state in which at least one of the image capturing range and the image capturing direction is adjusted.

Disclosed herein are an apparatus and method for classifying clothing attributes based on deep learning. The apparatus includes memory for storing at least one program and a processor for executing the program, wherein the program includes a first classification unit for outputting a first classification result for one or more attributes of clothing worn by a person included in an input image, a mask generation unit for outputting a mask tensor in which multiple mask layers respectively corresponding to principal part regions obtained by segmenting a body of the person included in the input image are stacked, a second classification unit for outputting a second classification result for the one or more attributes of the clothing by applying the mask tensor, and a final classification unit for determining and outputting a final classification result for the input image based on the first classification result and the second classification result.

An approach is provided for identifying partitions associated with erratic pedestrian behaviors and their correlations to points of interest. For example, the approach involves receiving sensor data associated with a geographic area. The approach also involves based on the sensor data, determining pedestrian-behavior parameter(s) respectively for partition(s). Each respective partition of the partition(s) represents a respective subarea of the geographic area, a respective time period, or a combination thereof. The approach further involves identifying at least one erratic partition from the partition(s) based on determining that a respective pedestrian-behavior parameter associated with the at least one erratic partition deviates from a baseline pedestrian-behavior parameter by at least a threshold extent. The approach further involves determining a correlation of the at least one erratic partition to at least one map feature of a geographic database. The approach further involves providing the correlation as an output.

Image analytics systems, methods, and computer program products to autonomously analyze an image to identify and detect features in the image, such as the horizon, and/or identify and detect objects of interest therein, such as, smoke or possible smoke. The image is captured, for example, by RGB cameras, and depicts a scene to be analyzed. The intelligent image analytic system is configured to provide alerts and/or other information to one or more concerned parties and/or computing systems to take an appropriate response.