Abstract: A navigating device includes: a camera configured to capture images within a field of view, the field of view depicting a scene including humans; a feature module configured to generate feature vectors based on the scene of humans in the images and to specify latent vectors that summarize movement of the humans in the scene based only on the camera images, the feature vectors summarizing the movement of the humans in the scene, and the latent vectors capturing a latent representation of trajectories in the scene of humans; a policy module configured to generate actions to be taken by the navigating device to navigate the scene of humans based on the feature vectors; and a propulsion control module configured to control one or more propulsion devices of the navigating device based on the actions to be taken generated by the policy module to navigate the scene of humans

Abstract: A training system for training a trained model for use by a navigating robot to perform visual navigation includes memory including N base virtual training environments, each of the N base virtual training environments including a field of view at a location within an indoor space, where N is an integer greater than 1. A randomization module is configured to generate N varied virtual training environments based on the N base virtual training environments, respectively, by varying at least one characteristic of the respective N base virtual training environments. A training module is configured to train the trained model for use by the navigating robot to perform visual navigation based on a training set including: the N base virtual training environments; and the N varied virtual training environments.

Abstract: A system and method for predicting a sequence of actions employ a Gated End-to-End Memory Policy Network (GMemN2NP), which includes a sequence of hop(s). Supporting memories of the hops include memory cells generated from observations made at different times. A sequence of actions is predicted, based on input agent-specific variables. For each action, the model, at each hop, outputs an updated controller state which is used as input to the next hop or, for the terminal hop, for computing the respective action. Each hop includes a transform gate mechanism which is used to control the influence of output of the supporting memories on the updated controller state. For the second and subsequent hops, respective actions are predicted, after using any intervening observations to update the supporting memories. The model is learned, on a training set of observations, to optimize the cumulative reward of a sequence of two or more actions.

Abstract: A training system for training a trained model for use by a navigating robot to perform visual navigation includes memory including N base virtual training environments, each of the N base virtual training environments including a field of view at a location within an indoor space, where N is an integer greater than 1. A randomization module is configured to generate N varied virtual training environments based on the N base virtual training environments, respectively, by varying at least one characteristic of the respective N base virtual training environments. A training module is configured to train the trained model for use by the navigating robot to perform visual navigation based on a training set including: the N base virtual training environments; and the N varied virtual training environments.

Abstract: A proactive interaction system includes memory which stores a contextual model. The contextual model includes supporting memory storing a representation of each of a set of past observations. Each of the past observations having an observed reward for a respective user for a respective action selected from a set of candidate actions. The contextual model is configured for estimating a reward for each of a current set of candidate actions, based on the stored representations of past observations and a representation of a current user. Each candidate action is associated with a respective action representation. A contextual bandit selects one of the candidate actions, based on the estimated reward for each of the set of candidate actions, to optimize a cumulative reward over a sequence of candidate action selections. An act output component performs a user-detectable act based on the selected one of the candidate actions.

Abstract: A proactive interaction system includes memory which stores a contextual model. The contextual model includes supporting memory storing a representation of each of a set of past observations. Each of the past observations having an observed reward for a respective user for a respective action selected from a set of candidate actions. The contextual model is configured for estimating a reward for each of a current set of candidate actions, based on the stored representations of past observations and a representation of a current user. Each candidate action is associated with a respective action representation. A contextual bandit selects one of the candidate actions, based on the estimated reward for each of the set of candidate actions, to optimize a cumulative reward over a sequence of candidate action selections. An act output component performs a user-detectable act based on the selected one of the candidate actions.

Abstract: Data characterizing a system is received at an electronic processor. For example, parking event data from parking sensors of a parking facility is received. The electronic processor constructs a current state for the system (e.g. parking occupancy state of the parking facility) at a current time from the received data. State probabilities at a future time are computed (e.g. occupancy state probabilities are computed for the parking facility) using a continuous-time Markov chain model modified by multiplying the time input to the model by a random variable and scaling the state probabilities by an expectation of the random variable. In parking occupancy forecasting, parking guidance information is generated based at least on the computed occupancy state probabilities, and is transmitted to an electronic device other than the electronic processor (e.g. a parking recommendation transmitted to a vehicle navigation device, or a control signal transmitted to a “lot full” sign).

Abstract: A system and method for predicting a sequence of actions employ a Gated End-to-End Memory Policy Network (GMemN2NP), which includes a sequence of hop(s). Supporting memories of the hops include memory cells generated from observations made at different times. A sequence of actions is predicted, based on input agent-specific variables. For each action, the model, at each hop, outputs an updated controller state which is used as input to the next hop or, for the terminal hop, for computing the respective action. Each hop includes a transform gate mechanism which is used to control the influence of output of the supporting memories on the updated controller state. For the second and subsequent hops, respective actions are predicted, after using any intervening observations to update the supporting memories. The model is learned, on a training set of observations, to optimize the cumulative reward of a sequence of two or more actions.

Abstract: A method for estimating a receiver's location in a wireless communication environment having several channels. Each channel has at least one signal parameter that varies with location differently from the other channels. A set of calibration data is determined for each calibration point, each set including the location and at least one measured signal parameter for each of several channels. The calibration data serve as a basis for a statistical model of the signal parameters versus a receiver's location. A set of observed signal parameters is determined, the set including at least one signal parameter for each of several channels at the receiver's location. A location estimate approximating the location of the receiver is determined on the basis of the statistical model and the set of observed signal parameters.

Abstract: A method for estimating a receiver's location (X) in a wireless communication environment (RN) having several channels. Each channel has at least one signal parameter (V) that varies with location (X) differently from the other channels. A set of calibration data (CD) is determined for each calibration point, each set comprising the location (X) and at least one measured signal parameter (V) for each of several channels. The calibration data (CD) serve as a basis for a statistical model (SM) of the signal parameters (V) versus a receiver's location. A set of observed signal parameters (CO) is determined, the set comprising at least one signal parameter (V) for each of several channels at the receiver's location (X). A location estimate (LE) approximating the location (X) of the receiver (R) is determined on the basis of the statistical model (SM) and the set of observed signal parameters (CO).