Visual-inertial fusion is crucial for a large amount of intelligent and autonomous applications, such as robot navigation and augmented reality. To bootstrap and achieve optimal state estimation, the spatial-temporal displacements between IMU and cameras must be calibrated in advance. Most existing calibration methods adopt continuous-time state representation, more specifically the B-spline. Despite these methods achieve precise spatial-temporal calibration, they suffer from high computational cost caused by continuous-time state representation. To this end, we propose a novel and extremely efficient calibration method that unleashes the power of discrete-time state representation. Moreover, the weakness of discrete-time state representation in temporal calibration is tackled in this paper. With the increasing production of drones, cellphones and other visual-inertial platforms, if one million devices need calibration around the world, saving one minute for the calibration of each device means saving 2083 work days in total. To benefit both the research and industry communities, our code will be open-source.

Deep reinforcement learning, a subfield of machine learning that combines reinforcement learning and deep learning, takes what's known as a reward function and learns to maximize the expected total reward. This works remarkably well, enabling systems to figure out how to solve Rubik's Cubes, beat world champions at chess, and more. But existing algorithms have a problem: They implicitly assume access to a perfect specification. In reality, tasks don't come prepackaged with rewards -- those rewards come from imperfect human reward designers. And it can be difficult to translate conceptual preferences into reward functions environments can calculate. To solve this problem, researchers at DeepMind and the University of California, Berkeley, have launched a competition called BASALT, where the goal of an AI system must be communicated through demonstrations, preferences, or some other form of human feedback.

As a part of the tech in Techstory, we bring you the latest in the technology from around the world under "TECH THIS WEEK!" every Sunday! This week we saw Mozilla's multi-process architecture finally cleared for takeoff and Apple's initiative Apple Energy to become an energy company. We also saw launching specs of much anticipated Bluetooth 5.0 and the announcement of AlphaGo's next duel with Go champion Ke Jie. In case, if you've missed any of those, along with how scientists have planned to reduce global warming, don't worry just keep reading! Mozilla's long-running project to bring multi-process architecture to hundreds of millions of Firefox users has finally met release criteria for a full-scale rollout. Nearly every other browser on the market has adopted multi-process architecture, splitting tabs and extensions into separate processes.

In retrospect, it seems crazy that we sent people to the moon with nothing there waiting for them. If something had gone wrong, there was no Plan B. We're probably not going to take a risk like that again, which is why we're working so hard on robots that can go to the moon or Mars to get things all set up and running and warm and cozy for us in advance. Setting up bases and habitats and doing exploring and whatnot may be the exciting extraterrestrial work, but there's other Very Important things that need to be done. One of the most important things is a high quality landing pad, and the Pacific International Space Center for Exploration (PISCES) has gotten a teleoperated robot to build one. The reason that you need a landing pad is that dust on the moon (and on Mars) is nasty, nasty stuff.