Skip to content
Only Research
Jan 23Tue
  1. 泡泡机器人SLAM · 微信公众号47

    Self-Supervised Depth and 3D Motion Estimation from Monocular Thermal Video Based on Multispectral Consistency Loss

    A study published at ICRA 2022 proposed a self-supervised learning method that simultaneously estimates depth and ego-motion from monocular thermal video. The method uses multispectral consistency consisting of a temperature consistency loss and a photometric consistency loss, and designs a differentiable forward warping module that transforms the thermal camera coordinate system into a visible-light camera, thereby providing a complementary self-supervised signal.

Jan 9Tue
  1. 泡泡机器人SLAM · 微信公众号41

    Robust Object Detection Under Adverse Conditions via Event Camera and RGB Camera Fusion

    An ICRA 2022 paper proposes a redundant sensor fusion model of event cameras and RGB cameras for robust object detection under image corruption and different weather conditions. The method takes voxel grid representations of events as input and processes frames and events separately through two parallel feature extractor networks. After training and evaluation on the public DSEC dataset, the fusion method improved robustness to common image corruptions by more than 30% compared with frame-only detection, and outperformed event-only detection.

Jan 2Tue
  1. 泡泡机器人SLAM · 微信公众号30

    [Bubble One Minute] A Compact and Flexible Angular-Spoke Wheel Robot for Uneven and Granular Terrain

    An ICRA 2022 paper proposes a mobile robot named DODO, which uses an angled-spoke wheel (ASW) design to improve driving speed on granular media, asphalt ground, and the transition zones between them. The study used the Taguchi method and an L9 orthogonal array to optimize design parameters, and through experiments selected a single-contact foot configuration, with front and rear ASW toe-in/toe-out angles of 15° and a 30° phase difference to reduce reciprocating motion of the center of mass along the Z axis.

Dec 25Mon
  1. 泡泡机器人SLAM · 微信公众号57

    NeRF-LOAM: Neural Implicit Representation for Large-Scale Incremental LiDAR Odometry and Mapping

    Bubble Turing Think Tank compiled and introduced the arXiv 2023 paper NeRF-LOAM, which the authors describe as the first neural implicit odometry and mapping system for large-scale environments using LiDAR data. It consists of three parts—Neural Tracking, Neural Mapping, and Mesh Reconstruction—all based on the Neural SDF they propose.

Dec 13Wed
Dec 11Mon
  1. 泡泡机器人SLAM · 微信公众号43

    [Bubble One Minute] Interval-based visual-inertial LiDAR SLAM using anchor poses (i-VIL SLAM)

    i-VIL SLAM proposes a visual-inertial LiDAR SLAM method that assumes only bounded sensor errors, propagates errors to the map and trajectory through interval analysis, and stabilizes uncertainty accumulation using anchor poses derived from GNSS/INS data. Experiments show that in the worst case, the 99.9% confidence ellipsoid derived from marginal covariance in classical SLAM graph optimization contains less than 64% of ground truth, whereas the pose interval bounds of this method contain ground truth for more than 96% of frames.

Mar 4Sat
  1. 机器人库 · 微信公众号55

    Flexible robot F3DB enables in vivo 3D bioprinting

    A team from the Medical Robotics Laboratory at the University of New South Wales has developed F3DB, a miniature flexible 3D bioprinter that can be inserted into the human body like an endoscope to deliver multiple layers of biological materials directly to internal organs and tissue surfaces; the research was published in the journal Advanced Materials. The device features a flexible rotating head and a snake-like robotic arm, has been tested in an artificial colon and printed materials in different shapes onto the surface of pig kidneys, and its smallest prototype has a diameter of about 11-13 micrometers; the system has been patented, and the next stage will be testing in live animals.

Feb 24Fri
  1. 机器人库 · 微信公众号38

    Study predicts nearly 40% of household chores may be handled by robots in 10 years

    Researchers from the University of Oxford in the UK and Ochanomizu University in Japan interviewed 65 AI experts and predicted that in 10 years about 39% of household chores may be completed by automated programs, with time spent buying daily necessities possibly cut by nearly 60% and house cleaning saving 46% of time. The report also noted that only 28% of caregiving work such as education, accompanying children, and caring for the elderly can be automated, and highly automated households may face privacy issues.

Oct 12Wed
Sep 20Tue
  1. 机器人产业网 · 微信公众号29

    Brief: Bubble microrobots achieve remote speed modulation; study on timing in human-robot collaborative operation; multiple institutions publish a review of efficient NLP methods, and more

    A team from the Institute of Mechanics, Chinese Academy of Sciences, published research in Small on magnetically controlled bubble microrobots. By adjusting the orientation of Janus microspheres through a 3D Helmholtz coil magnetic control system, they achieved remote speed modulation, and used microbubble collapse jets to accomplish propulsion, anchoring, and grasping; surface capillary waves can sweep particles in the far field.