Abstract: Described is a system for improving generalization of an agent, such as an autonomous vehicle, to unanticipated environmental changes. A set of concepts from the agent's experiences of an environment are extracted and consolidated into an episodic world model. Using the episodic world model, a dream sequence of prospective simulations, based on a selected set of concepts and constrained by the environment's semantics and dynamics, is generated. The dream sequence is converted into a sensor data format, which is used for augmented training of the agent to operate in the environment with improved generalization to unanticipated changes in the environment.

Abstract: Described is a system for continual learning using experience replay. In operation, the system receives a plurality of tasks sequentially, from which a current task is fed to an encoder. The current task has data points associated with the current task. The encoder then maps the data points into an embedding space, which reflects the data points as discriminative features. A decoder then generates pseudo-data points from the discriminative features, which are provided back to the encoder. The discriminative features are updated in the embedding space based on the pseudo-data points. The encoder then learns (updates) a classification of a new task by matching the new task with the discriminative features in the embedding space.

Abstract: Described is a system for competency assessment of an autonomous system. The system extracts semantic concepts representing a situation. Actions taken by the autonomous system are associated with semantic concepts that are activated when the actions are taken in the situation. The system measures an outcome of the actions taken in the situation and generates a reward metric. The semantic concepts representing the situation are stored as a memory with the actions taken in the situation and the reward metric as a memory. A prospective simulation is generated based on recall of the memory. A competency metric and an experience metric are determined. Competent operational control of the autonomous system is maintained when at least one of the competency metric and the experience metric is above a minimum value. An alert is generated when at least one of the competency metric and the experience metric falls below the minimum value.

Abstract: Described is a system for decoding and validating memory consolidation. During operation, the system receives electroencephalographic (EEG) data while a subject is performing a specific task. Nuisance signals are then removed from the EEG data, resulting in a nuisance free signal. Skill feature vectors are generated from the nuisance free signal using time-invariant feature extraction. A skill classifier can then be trained for the specific task based on the skill feature vectors to generate a subject specific model regarding a memory replay for the specific task. Finally, electrodes in a neural cap are activated based on the memory replay.

Abstract: In an embodiment of the present invention, a method for generating a prediction of ability of a subject to perform a task in a future time step includes receiving performance data corresponding to a performance of the subject on the task; receiving a plurality of biometric inputs computed based on physiological data during the performance of the subject on the task; identifying a numerical relationship between the performance data and the plurality of biometric inputs; generating a modulation parameter for each of the plurality of biometric inputs based on the identified numerical relationship; loading a plurality of state variable inputs produced by a generic model of performance; and generate the prediction of ability to perform the task at the prediction time, generated by a trained performance predictor based on biometric inputs predicted based on the modulation parameters.

Abstract: A neuro-stimulation system having: a controller; a user interface arranged for, under control of the controller, providing a user with a series of information elements to be learned by the user; a non-invasive brain stimulator arranged for, under the control of the controller, tagging each information element of said series of information elements by stimulating a brain of the user with a different, unique, associated brain stimulus when the user is provided said information element by said user interface; and a non-invasive brain sensor arranged for sending to the controller data indicating a slow-wave sleep period of the user; wherein the controller is further arranged for cueing each information element of said series of information elements by causing the stimulator to stimulate the brain of the user with the brain stimuli associated with said series of information elements during said slow-wave sleep period of the user.

Abstract: A neuro-stimulation system comprising a user interface arranged for providing a user with a series of sensory stimuli, and for selecting, in response to a user input a group of selected sensory stimuli out of said series of sensory stimuli; a non-invasive brain stimulator arranged for stimulating the brain of the user with a different, unique, associated brain stimulus each time the user interface selects one stimulus of said group of selected sensory stimuli; and a controller arranged for detecting slow-wave sleep of the user and for causing the stimulator to stimulate the brain of the user with the brain stimuli associated with said group of selected sensory stimuli during the slow-wave sleep of the user.

Abstract: Described is a system for cueing a specific memory in a waking state. The system sends an initiation signal to a memory recall controller to select a stored stimulation pattern previously associated with a specific memory of an event. The system signals to a memory recall controller to initiate delivery of the selected stimulation pattern to a brain in a waking state for a duration of the event via a brain stimulation system. Following completion of the event, the system signals for the memory recall controller to stop the brain stimulation system from delivering the selected stimulation pattern.

Abstract: Described is an improved brain-machine interface including a neural interface and a controllable device in communication with the neural interface. The neural interface includes a neural device with one or more sensors for collecting signals of interest and one or more processors for conditioning the signals of interest, extracting salient neural features from and decoding the conditioned signals of interest, and generating a control command for the controllable device. The controllable device performs one or more operations according to the control command, and the neural device administers neuromodulation stimulation to reinforce operation of the controllable device.

Abstract: An autonomous navigation system for a vehicle includes a controller configured to control the vehicle, sensors configured to detect objects in a path of the vehicle, nonvolatile memory including an artificial neural network configured to classify the objects detected by the sensors, and a processor. The artificial neural network includes a series of neurons in each of an input layer, at least one hidden layer, and an output layer. The memory includes instructions which, when executed by the processor, cause the processor to train the artificial neural network on a first task, identify, utilizing a contrastive excitation backpropagation algorithm, important neurons for the first task, identify, utilizing a learning algorithm, important synapses between the neurons for the first task based on the important neurons identified, and rigidify the important synapses to achieve selective plasticity of the series of neurons in the artificial neural network.

Abstract: Described is a system for online vehicle recognition in an autonomous driving environment. Using a learning network comprising an unsupervised learning component and a supervised learning component, images of moving vehicles extracted from videos captured in the autonomous driving environment are learned and classified. Vehicle feature data is extracted from input moving vehicle images. The extracted vehicle feature data is clustered into different vehicle classes using the unsupervised learning component. Vehicle class labels for the different vehicle classes are generated using the supervised learning component. Based on a vehicle class label for a moving vehicle in the autonomous driving environment, the system selects an action to be performed by the autonomous vehicle, and causes the selected action to be performed by the autonomous vehicle in the autonomous driving environment.

Abstract: A system configured to identify a target in a synthetic aperture radar signal includes: a feature extractor configured to extract a plurality of features from the synthetic aperture radar signal; an input spiking neural network configured to encode the features as a first plurality of spiking signals; a multi-layer recurrent neural network configured to compute a second plurality of spiking signals based on the first plurality of spiking signals; a readout neural layer configured to compute a signal identifier based on the second plurality of spiking signals; and an output configured to output the signal identifier, the signal identifier identifying the target.

Abstract: A method for retraining an artificial neural network trained on data from an old task includes training the artificial neural network on data from a new task different than the old task, calculating a drift, utilizing Sliced Wasserstein Distance, in activation distributions of a series of hidden layer nodes during the training of the artificial neural network with the new task, calculating a number of additional nodes to add to at least one hidden layer based on the drift in the activation distributions, resetting connection weights between input layer nodes, hidden layer nodes, and output layer nodes to values before the training of the artificial neural network on the data from the new task, adding the additional nodes to the at least one hidden layer, and training the artificial neural network on data from the new task.

Abstract: Described is a system for weakening an undesirable memory. The system initiates application of a first pattern of spatiotemporally distributed transcranial stimulation via a set of electrodes to a subject who is in a calm mental state, causing association of the first pattern of spatiotemporally distributed transcranial stimulation with the calm mental state. The system then initiates application of the first pattern of spatiotemporally distributed transcranial stimulation via the set of electrodes when the undesirable memory is recalled by the subject, causing recall of the calm mental state and reconsolidation of the undesirable memory with the calm mental state.

Abstract: Described is a system for continual adaptation of a machine learning model implemented in an autonomous platform. The system adapts knowledge previously learned by the machine learning model for performance in a new domain. The system receives a consecutive sequence of new domains comprising new task data. The new task data and past learned tasks are forced to share a data distribution in an embedding space, resulting in a shared generative data distribution. The shared generative data distribution is used to generate a set of pseudo-data points for the past learned tasks. Each new domain is learned using both the set of pseudo-data points and the new task data. The machine learning model is updated using both the set of pseudo-data points and the new task data.

Abstract: A brain-machine interface system configured to decode neural signals to control a target device includes a sensor to sample the neural signals, and a computer-readable storage medium having software instructions, which, when executed by a processor, cause the processor to transform the neural signals into a common representational space stored in the system, provide the common representational space as a state representation to inform an Actor recurrent neural network policy of the system, generate and evaluate, utilizing a deep recurrent neural network of the system having a generative sequence decoder, predictive sequences of control signals, supply a control signal to the target device to achieve an output of the target device, determine an intrinsic biometric-based reward signal, from the common representational space, based on an expectation of the output of the target device, and supply the intrinsic biometric-based reward signal to a Critic model of the system.

Abstract: A system configured to identify a target in a synthetic aperture radar signal includes: a feature extractor configured to extract a plurality of features from the synthetic aperture radar signal; a spiking neural network configured to encode the features as a plurality of spiking signals; a readout neural layer configured to compute a signal identifier based on the spiking signals; and an output configured to output the signal identifier, the signal identifier identifying the target.

Abstract: Described is a system for competency assessment of an autonomous system. The system extracts semantic concepts representing a situation. Actions taken by the autonomous system are associated with semantic concepts that are activated when the actions are taken in the situation. The system measures an outcome of the actions taken in the situation and generates a reward metric. The semantic concepts representing the situation are stored as a memory with the actions taken in the situation and the reward metric as a memory. A prospective simulation is generated based on recall of the memory. A competency metric and an experience metric are determined. Competent operational control of the autonomous system is maintained when at least one of the competency metric and the experience metric is above a minimum value. An alert is generated when at least one of the competency metric and the experience metric falls below the minimum value.

Abstract: Described is a system for adaptable neurostimulation intervention. The system monitors a set of neurophysiological signals in real-time and updates a physiological and behavioral model. The set of neurophysiological signals are classified in real-time based on the physiological and behavioral model. A neurostimulation intervention schedule is generated based on the classified set of neurophysiological signals. The system activates electrodes via a neurostimulation intervention system to cause a timed neurostimulation intervention to be administered based on the neurostimulation intervention schedule. The neurostimulation intervention schedule and timed neurostimulation intervention are refined based on new sets of neurophysiological signals.

Abstract: Described is a system for continual learning using experience replay. In operation, the system receives a plurality of tasks sequentially, from which a current task is fed to an encoder. The current task has data points associated with the current task. The encoder then maps the data points into an embedding space, which reflects the data points as discriminative features. A decoder then generates pseudo-data points from the discriminative features, which are provided back to the encoder. The discriminative features are updated in the embedding space based on the pseudo-data points. The encoder then learns (updates) a classification of a new task by matching the new task with the discriminative features in the embedding space.