Abstract: A system for estimating a current camera pose corresponding to a current point in time using a previous camera pose corresponding to a previous point in time, of a camera configured to generate a sequence of image frames. The system performs operations, including: generating, using one or more neural networks, a neural network pose prediction for the current image frame; and adjusting a previous camera pose using inertial measurement unit data representing a motion of the camera between the previous point in time and the current point in time, to provide an inertial measurement unit pose prediction for the current point in time. The inertial measurement unit pose prediction, and the neural network pose prediction are combined in order to estimate the current camera pose.

Abstract: Disclosed is a memory structure including an array of memory cells and a read circuit. The read circuit includes two registers configured to capture and store two different digital-to-analog converter (DAC) codes, which correspond to two different reference currents that approximate two different output currents generated on a bitline during consecutive single-ended current sensing processes directed to the same selected memory cell but using different input voltages. Optionally, the read circuit can also include a current-voltage (I-V) slope calculator, which uses the two different DAC codes to calculate an I-V slope characteristic of the selected memory cell, and a bit generator, which performs a comparison of the I-V slope characteristic and a reference I-V slope characteristic and based on results of the comparison, generates and outputs a bit with a logic value that represents the data storage state of the selected memory cell. Also disclosed is an associated method.

Abstract: A system for estimating a current camera pose corresponding to a current point in time using a previous camera pose corresponding to a previous point in time, of a camera configured to generate a sequence of image frames. The system performs operations, including: generating, using one or more neural networks, a neural network pose prediction for the current image frame; and adjusting a previous camera pose using inertial measurement unit data representing a motion of the camera between the previous point in time and the current point in time, to provide an inertial measurement unit pose prediction for the current point in time. The inertial measurement unit pose prediction, and the neural network pose prediction are combined in order to estimate the current camera pose.