Disclosed herein are systems, methods, and devices for classifying ophthalmic images according to disease type, state, and stage. The disclosed invention details systems, methods, and devices to perform the aforementioned classification based on weighted-linkage of an ensemble of machine learning models. In some parts, each model is trained on a training data set and tested on a test dataset. In other parts, the models are ranked based on classification performance, and model weights are assigned based on model rank. To classify an ophthalmic image, that image is presented to each model of the ensemble for classification, yielding a probabilistic classification score—of each model. Using the model weights, a weighted-average of the individual model-generated probabilistic scores is computed and used for the classification.

Mechanisms are provided to provide an improved computer tool for determining and mitigating the presence of adversarial inputs to an image classification computing model. A machine learning computer model processes input data representing a first image to generate a first classification output. A cohort of second image(s), that are visually similar to the first image, is generated based on a comparison of visual characteristics of the first image to visual characteristics of images in an image repository. A cohort-based machine learning computer model processes the cohort of second image(s) to generate a second classification output and the first classification output is compared to the second classification output to determine if the first image is an adversarial image. In response to the first image being determined to be an adversarial image, a mitigation operation by a mitigation system is initiated.

Techniques for determining value of a home by applying one or more neural network models to images of spaces in the home. The techniques include: obtaining at least one image of a first space inside or outside of a home; determining a type of the first space by processing the at least one image of the first space with a first neural network model; identifying at least one feature in the first space by processing the at least one image with a second neural network model different from the first neural network model and trained using images of spaces of a same type as the first space; and determining a value of the home at least in part by using the at least one feature as input to a machine learning model different from the first neural network model and the second neural network model.

A method and device for automatically classifying document hardcopy images by using document hardcopy image descriptors are provided. The method and device include providing a document hardcopy image, the document hardcopy image having image features, extracting image descriptors by a first set of image descriptor extractors, each image descriptor of the image descriptors being descriptive of the image features of the document hardcopy image, estimating class probabilities of the document hardcopy image by multiple trained classifiers based on the image descriptors, determining a most probable class of the document hardcopy image by a trained meta-classifier based on the class probabilities estimated by the multiple trained classifiers, inputting the document hardcopy image and the most probable class of the document hardcopy image to an assigner, and assigning, by the assigner, the most probable class determined by the trained meta-classifier to the document hardcopy image to obtain a classified document hardcopy image.

Methods and systems are described for generating dynamic conversational responses using two-tier machine learning models. The dynamic conversational responses may be generated in real time and reflect the likely goals and/or intents of a user. The two-tier machine learning model may include a first tier that determines an intent cluster based on a feature input, and a second tier that determines a specific intent from the cluster.

An annotation system uses annotations for a first set of sensor measurements from a first sensor to identify annotations for a second set of sensor measurements from a second sensor. The annotation system identifies reference annotations in the first set of sensor measurements that indicates a location of a characteristic object in the two-dimensional space. The annotation system determines a spatial region in the three-dimensional space of the second set of sensor measurements that corresponds to a portion of the scene represented in the annotation of the first set of sensor measurements. The annotation system determines annotations within the spatial region of the second set of sensor measurements that indicates a location of the characteristic object in the three-dimensional space.

Aspects of the subject matter disclosed herein include methods, systems, and other techniques for training, in a first phase, an object classifier neural network with a first set of training data, the first set of training data including a first plurality of training examples, each training example in the first set of training data being labeled with a coarse-object classification; and training, in a second phase after completion of the first phase, the object classifier neural network with a second set of training data, the second set of training data including a second plurality of training examples, each training example in the second set of training data being labeled with a fine-object classification.

Method and apparatus for object detection by a robot are provided. The method comprises analyzing using a set of trained detection models, one or more first images of an environment of the robot to detect one or more objects in the environment of the robot, generating at least one fine-tuned model by training one or more of the trained detection models in the set, wherein the training is based on a second image of the environment of the robot and annotations associated with the second image, wherein the annotations identify one or more objects in the second image, updating the set of trained detection models to include the generated at least one fine-tuned model, and analyzing using the updated set of trained detection models, one or more third images of the environment of the robot to detect one or more objects in the environment.

A radar anti-spoofing system for an autonomous vehicle includes a plurality of radar sensors that generate a plurality of input detection points representing radio frequency (RF) signals reflected from objects and a controller in electronic communication with the plurality of radar sensors. The controller executes instructions to determine time-matched clusters that represent objects located in an environment surrounding the autonomous vehicle based on the input detection points from the plurality of radar sensors. The controller determines an adjusted signal to noise (SNR) measure for a specific time-matched cluster by dividing an SNR of the specific time-matched cluster by a range measurement of the specific time-matched cluster. The controller determines a velocity-ratio measure of the time-matched cluster by dividing a motion-based velocity by a Doppler-frequency velocity, and identifies the time-matched cluster as either a ghost object or a real object.

A system for event detection and reporting has primary sensors for producing raw data from observing the proximate area, one or more processors for the primary sensors, for processing the raw data to produce output, a centralized controller to which each of the processors is connected, for receiving the output, a security network for communicating between the sensors, the controller, and one or more remote terminals, the security network having an alarm, and a plurality of auxiliary sensors for providing secondary sensor information to the one or more processors wherein the controller provides an alarm to the remote terminals through the security network if an event has occurred. A method of event detection has the steps of receiving a control command, alarm message and sensor information, determining if an event has occurred, and sending alarm messages and filtered sensor information to remote terminals if an event has occurred.