Abstract: Devices, systems, and methods for machine learning (ML) automatic target recognition (ATR) decision explanation are provided. A method can include receiving an object specification matrix from an object model database that indicates, for each of a plurality of physical portions of an object, whether each of a plurality of features are present or absent in a physical portion of the physical portions of the object and a proportional physical displacement between the features in the object, receiving feature data indicating for an image of a portion of the object, a likelihood whether each of features are present in the image, determining based on the object specification matrix and the feature data, a probability and corresponding uncertainty that the image corresponds to the object, and providing the probability and corresponding uncertainty of the object to help classify the object.

Abstract: An automatic target recognizer system including: a database that stores target recognition data including multiple reference features associated with each of multiple reference targets; a pre-selector that selects a portion of the target recognition data based on a reference gating feature of the multiple reference features; a preprocessor that processes an image received from an image acquisition system which is associated with an acquired target and determines an acquired gating feature of the acquired target; a feature extractor and processor that discriminates the acquired gating feature with the reference gating feature and, if there is a match, extracts multiple segments of the image and detects the presence, absence, probability or likelihood of one of multiple features of each of the multiple reference targets; a classifier that generates a classification decision report based on a determined classification of the acquired target; and a user interface that displays the classification decision report.

Abstract: Devices, systems, and methods for removing non-linearities in an inverse synthetic aperture radar (ISAR) image are provided. A method includes estimating pitch and roll about range and doppler axes of a time series of ISAR images including the ISAR image, interpolating ISAR image data based on the estimated pitch and roll resulting in interpolated ISAR image data, and resampling based on the interpolated ISAR image data and the time series of TSAR images resulting in an enhanced image.

Abstract: An automatic target recognizer system including: a database that stores target recognition data including multiple reference features associated with each of multiple reference targets; a pre-selector that selects a portion of the target recognition data based on a reference gating feature of the multiple reference features; a preprocessor that processes an image received from an image acquisition system which is associated with an acquired target and determines an acquired gating feature of the acquired target; a feature extractor and processor that discriminates the acquired gating feature with the reference gating feature and, if there is a match, extracts multiple segments of the image and detects the presence, absence, probability or likelihood of one of multiple features of each of the multiple reference targets; a classifier that generates a classification decision report based on a determined classification of the acquired target; and a user interface that displays the classification decision report.

Abstract: Devices, systems, and methods for machine learning model generation. A method can include generating image chips of an image. The image chips can each provide a view of a different extent of an object in the image. Based on an object definition that indicates respective features of the object and a location of the respective features along a length of the object, it can be determined whether any of the image chips include any of the respective features. Each image chip of the image chips can be labeled to include an indication of any of the features included in the image chip resulting in labelled image chips. The method can include training an ensemble classifier based on the labelled image chips resulting in a trained ensemble classifier.

Abstract: Processing circuitry for a deep neural network can include input/output ports, and a plurality of neural network layers coupled in order from a first layer to a last layer, each of the plurality of neural network layers including a plurality of weighted computational units having circuitry to interleave forward propagation of computational unit input values from the first layer to the last layer and backward propagation of output error values from the last layer to the first layer.

Abstract: An automatic target recognizer system including: a database that stores target recognition data including multiple reference features associated with each of multiple reference targets; a pre-selector that selects a portion of the target recognition data based on a reference gating feature of the multiple reference features; a preprocessor that processes an image received from an image acquisition system which is associated with an acquired target and determines an acquired gating feature of the acquired target; a feature extractor and processor that discriminates the acquired gating feature with the reference gating feature and, if there is a match, extracts multiple segments of the image and detects the presence, absence, probability or likelihood of one of multiple features of each of the multiple reference targets; a classifier that generates a classification decision report based on a determined classification of the acquired target; and a user interface that displays the classification decision report.

Abstract: An automatic target recognizer system including: a database that stores target recognition data including multiple reference features associated with each of multiple reference targets; a pre-selector that selects a portion of the target recognition data based on a reference gating feature of the multiple reference features; a preprocessor that processes an image received from an image acquisition system which is associated with an acquired target and determines an acquired gating feature of the acquired target; a feature extractor and processor that discriminates the acquired gating feature with the reference gating feature and, if there is a match, extracts multiple segments of the image and detects the presence, absence, probability or likelihood of one of multiple features of each of the multiple reference targets; a classifier that generates a classification decision report based on a determined classification of the acquired target; and a user interface that displays the classification decision report.

Abstract: A dynamically adaptive neural network processing system includes memory to store instructions representing a neural network in contiguous blocks, hardware acceleration (HA) circuitry to execute the neural network, direct memory access (DMA) circuitry to transfer the instructions from the contiguous blocks of the memory to the HA circuitry, and a central processing unit (CPU) to dynamically modify a linked list representing the neural network during execution of the neural network by the HA circuitry to perform machine learning, and to generate the instructions in the contiguous blocks of the memory based on the linked list.

Abstract: Processing circuitry for a deep neural network can include input/output ports, and a plurality of neural network layers coupled in order from a first layer to a last layer, each of the plurality of neural network layers including a plurality of weighted computational units having circuitry to interleave forward propagation of computational unit input values from the first layer to the last layer and backward propagation of output error values from the last layer to the first layer.

Abstract: A dynamically adaptive neural network processing system includes memory to store instructions representing a neural network in contiguous blocks, hardware acceleration (HA) circuitry to execute the neural network, direct memory access (DMA) circuitry to transfer the instructions from the contiguous blocks of the memory to the HA circuitry, and a central processing unit (CPU) to dynamically modify a linked list representing the neural network during execution of the neural network by the HA circuitry to perform machine learning, and to generate the instructions in the contiguous blocks of the memory based on the linked list.

Abstract: A vehicle comprises a body, a first leg, and a second leg. The first leg has a proximal end jointed to the body, a distal end, and a first foot located on the distal end. A first maximum working envelope is associated with the first foot, where the first working range is inscribed within a first maximum working envelope. Likewise, the second leg has a proximal end jointed to the body in-line with the first leg, a distal end, and a second foot located on the distal end. A second maximum working envelope is associated with the second foot, where the second working range is inscribed within a second maximum working envelope. The vehicle thus defines a single-track multi-legged vehicle where the first leg and the second leg are attached to the body one behind the other, substantially parallel to a major axis of motion of the vehicle.

Abstract: A vehicle comprises a body, a first leg, and a second leg. The first leg has a proximal end jointed to the body, a distal end, and a first foot located on the distal end. A first maximum working envelope is associated with the first foot, where the first working range is inscribed within a first maximum working envelope. Likewise, the second leg has a proximal end jointed to the body in-line with the first leg, a distal end, and a second foot located on the distal end. A second maximum working envelope is associated with the second foot, where the second working range is inscribed within a second maximum working envelope. The vehicle thus defines a single-track multi-legged vehicle where the first leg and the second leg are attached to the body one behind the other, substantially parallel to a major axis of motion of the vehicle.

Abstract: A single-track legged vehicle includes a frame having a length corresponding to a forward/backward direction of travel, and a width that is normal to the length. The single-track legged vehicle also includes a plurality of jointed legs attached to the frame, where the plurality of jointed legs are aligned in-line, one behind the other along the length of the frame. Moreover, a control system is in communication with mechanisms that operate the plurality of jointed legs to control the legged vehicle for movement by decoupling the positioning of each of the plurality of jointed legs along a major axis of motion corresponding to the length of the frame from the positioning of each of the plurality of jointed legs along the width of the body, normal to major axis of motion. Yet further, a power source that powers the control system and mechanisms that operate the plurality of jointed legs.

Abstract: A single-track legged vehicle includes a frame having a length corresponding to a forward/backward direction of travel, and a width that is normal to the length. The single-track legged vehicle also includes a plurality of jointed legs attached to the frame, where the plurality of jointed legs are aligned in-line, one behind the other along the length of the frame. Moreover, a control system is in communication with mechanisms that operate the plurality of jointed legs to control the legged vehicle for movement by decoupling the positioning of each of the plurality of jointed legs along a major axis of motion corresponding to the length of the frame from the positioning of each of the plurality of jointed legs along the width of the body, normal to major axis of motion. Yet further, a power source that powers the control system and mechanisms that operate the plurality of jointed legs.

Abstract: A single track legged vehicle having a body and at least three in-line legs aligned one behind the other is operated by controlling each in-line leg to develop a desired ballistic flight trajectory, by controlling foot force and torque during a first phase that produces thrust; controlling foot movement during a second phase that transitions from thrusting to flight, controlling in-line leg movement during a third phase characterized by flight; controlling foot positioning during a fourth phase characterized by a transition from flight to landing; and controlling foot force and torque during a fifth phase characterized by landing of the foot of the corresponding in-line leg. Each in-line leg is transitioned through the first phase, the second phase the third phase, the fourth phase and the fifth phase to propel and torque the body along three axes according to the desired ballistic flight trajectory.

Abstract: A temporal controller for mobile robot path planning includes a sensor module for receiving data corresponding to spatial locations of at least one object, and a temporal control module operatively coupled to the sensor module, the temporal control module configured to predict future locations of the at least one object based on data received by the sensor module. The controller further includes a temporal simulation module operatively coupled to the temporal control module, wherein the temporal simulation module configured to use the predicted future locations of the at least one object to simulate multiple robot motion hypothesis for object avoidance and trajectory planning.

Abstract: A single track legged vehicle having a body and at least three in-line legs aligned one behind the other is operated by controlling each in-line leg to develop a desired ballistic flight trajectory, by controlling foot force and torque during a first phase that produces thrust; controlling foot movement during a second phase that transitions from thrusting to flight, controlling in-line leg movement during a third phase characterized by flight; controlling foot positioning during a fourth phase characterized by a transition from flight to landing; and controlling foot force and torque during a fifth phase characterized by landing of the foot of the corresponding in-line leg. Each in-line leg is transitioned through the first phase, the second phase the third phase, the fourth phase and the fifth phase to propel and torque the body along three axes according to the desired ballistic flight trajectory.

Abstract: A single track in-line legged vehicle is controlled to coordinate movement along a desired single-track trajectory by causing each in-line leg to selectively perform a stance-to-flight phase, a flight phase, a flight-to-stance phase, and a stance phase. During the stance-to-flight phase, reaction forces and torques between a foot and the ground are unloaded to lift the foot off the ground. During the flight phase, a foot moves in the same general direction and at a generally faster rate as a major direction of motion of the vehicle body. During the flight-to-stance phase, foot positioning is controlled to place a foot on the ground according to the desired single-track trajectory. During the stance phase, foot-to-ground interaction develops reaction forces and torques that are transferred from the foot through the corresponding in-line leg to propel, torque, and stabilize the body in the x, y, z, pitch, roll, and yaw axes.

Abstract: A single track in-line legged vehicle is controlled to coordinate movement along a desired single-track trajectory by causing each in-line leg to selectively perform a stance-to-flight phase, a flight phase, a flight-to-stance phase, and a stance phase. During the stance-to-flight phase, reaction forces and torques between a foot and the ground are unloaded to lift the foot off the ground. During the flight phase, a foot moves in the same general direction and at a generally faster rate as a major direction of motion of the vehicle body. During the flight-to-stance phase, foot positioning is controlled to place a foot on the ground according to the desired single-track trajectory. During the stance phase, foot-to-ground interaction develops reaction forces and torques that are transferred from the foot through the corresponding in-line leg to propel, torque, and stabilize the body in the x, y, z, pitch, roll, and yaw axes.