The field of 3D perception and localization is rapidly advancing, with a focus on improving the accuracy and robustness of various sensors and algorithms. Recent developments have seen the introduction of new datasets and benchmarks, such as Bench-RNR and MCOD, which aim to evaluate the performance of different LiDAR scanning patterns and camouflaged object detection methods. Additionally, innovative approaches like Omni-LIVO and CrossI2P have been proposed to enhance the accuracy and efficiency of visual-inertial-LiDAR odometry and image-to-point cloud registration. Noteworthy papers include Bench-RNR, which provides a comprehensive dataset for benchmarking repetitive and non-repetitive scanning LiDARs, and Omni-LIVO, which introduces a tightly coupled multi-camera LIVO system for robust and accurate odometry. Other notable papers include MCOD, which presents a challenging benchmark for multispectral camouflaged object detection, and CrossI2P, which proposes a self-supervised framework for image-to-point cloud registration.
Advances in 3D Perception and Localization
Sources
Bench-RNR: Dataset for Benchmarking Repetitive and Non-repetitive Scanning LiDAR for Infrastructure-based Vehicle Localization
Omni-LIVO: Robust RGB-Colored Multi-Camera Visual-Inertial-LiDAR Odometry via Photometric Migration and ESIKF Fusion
MCOD: The First Challenging Benchmark for Multispectral Camouflaged Object Detection
Self-Supervised Cross-Modal Learning for Image-to-Point Cloud Registration
Global Regulation and Excitation via Attention Tuning for Stereo Matching
Towards Sharper Object Boundaries in Self-Supervised Depth Estimation
OrthoLoC: UAV 6-DoF Localization and Calibration Using Orthographic Geodata
TriFusion-AE: Language-Guided Depth and LiDAR Fusion for Robust Point Cloud Processing
RS3DBench: A Comprehensive Benchmark for 3D Spatial Perception in Remote Sensing
LiDAR Point Cloud Image-based Generation Using Denoising Diffusion Probabilistic Models
Towards Robust LiDAR Localization: Deep Learning-based Uncertainty Estimation
RoSe: Robust Self-supervised Stereo Matching under Adverse Weather Conditions
CU-Multi: A Dataset for Multi-Robot Collaborative Perception
Lidar-based Tracking of Traffic Participants with Sensor Nodes in Existing Urban Infrastructure
Optical Ocean Recipes: Creating Realistic Datasets to Facilitate Underwater Vision Research
Techno-Economic analysis for Smart Hangar inspection operations through Sensing and Localisation at scale