RGB-D Visual Simultaneous Localization and Mapping (SLAM) Application
PulseAugur coverage of RGB-D Visual Simultaneous Localization and Mapping (SLAM) Application — every cluster mentioning RGB-D Visual Simultaneous Localization and Mapping (SLAM) Application across labs, papers, and developer communities, ranked by signal.
- instance of CatalyzeX 90%
- instance of ScienceCast 70%
- instance of Gotit.pub 70%
- used by RGB color model 70%
- instance of RGB color model 70%
- developed ScienceCast 70%
- developed CatalyzeX 70%
- developed Gotit.pub 70%
- used by lidar 70%
- used by SCANNET 70%
- uses 3D computer graphics 70%
- used by 3D computer graphics 60%
10 day(s) with sentiment data
-
New ObjectTwin protocol measures accuracy of generated 3D assets in captured scenes
Researchers have introduced ObjectTwin, a novel measurement protocol designed to assess the accuracy of generated 3D assets when integrated into captured scenes. This protocol quantifies the "asset correspondence gap," …
-
New Bi-Mamba Model Enhances Robot Navigation with Real-Time Trajectory Prediction
Researchers have developed TrajMamba, a novel bidirectional state space model designed for real-time human trajectory prediction using RGB-D cameras. This model addresses the train-deploy pipeline gap by aligning pose e…
-
AI system precisely locates cow anatomy in real-time milking parlors
Researchers have developed a new computer vision system capable of accurately identifying specific anatomical regions on dairy cows in real-time within commercial rotary milking parlors. This system utilizes RGB-D image…
-
New DAFIC framework uses depth data to improve few-shot image classification
Researchers have introduced DAFIC, a novel framework designed to enhance few-shot image classification by integrating monocular depth information. This approach addresses visual ambiguity and background interference by …
-
F2G-Pose: Real-time RGB-D object pose estimation framework unveiled
Researchers have introduced F2G-Pose, a novel framework designed for real-time, category-level object pose estimation from RGB-D images. This system bypasses the need for instance-specific CAD models or reference views,…
-
New CVG-SGM method enhances object identity persistence in 4D scene graphs
Researchers have developed a new method called CVG-SGM to improve the persistence of object identity in 4D scene graphs, particularly when objects become temporarily invisible. This system uses a novel memory mechanism …
-
New pipeline validates markerless spinal tracking using intraoperative X-rays
Researchers have developed a new pipeline to validate markerless intraoperative tracking of spinal anatomy using RGB-D sensors. This method integrates with existing C-arm fluoroscopy workflows, deriving per-vertebra pos…
-
Study questions dynamic-point filtering effectiveness in ORB-SLAM2
A new study published on arXiv investigates the effectiveness of dynamic-point filtering in feature-based visual SLAM systems, specifically examining ORB-SLAM2. Researchers created synthetic indoor scenes with varying l…
-
New models tackle novel view synthesis and camera pose estimation
Two new research papers introduce novel methods for synthesizing new views of a scene from existing images and estimating camera poses. The first paper, LVS, focuses on reusing previously rendered images to improve inte…
-
PointVGGT framework offers zero-shot multiview RGB-D point cloud registration
Researchers have introduced PointVGGT, a novel zero-shot framework for multiview RGB-D point cloud registration. This method bypasses traditional pairwise registration by leveraging visual geometry foundation models to …
-
New datasets and frameworks advance visual-tactile learning for AI
Researchers have introduced two new datasets and frameworks aimed at advancing visual-tactile learning for embodied AI. TouchScale provides 500 hours of synchronized visual and tactile data from human interactions, sign…
-
TouchScale dataset offers 500 hours of human vision and touch data for AI learning
Researchers have introduced TouchScale, a new dataset comprising 500 hours of human vision and touch data specifically designed for visual-tactile learning. This dataset aims to address the scarcity of synchronized tact…
-
New CrowdOcc framework enhances robot scene completion in crowded indoor spaces
Researchers have introduced CrowdOcc, a new framework and dataset designed to improve monocular semantic scene completion for quadruped robots operating in crowded indoor environments. The system addresses challenges po…
-
RYOPO system enables real-time object pose estimation
Researchers have developed RYOPO, a novel end-to-end trainable system for category-level object pose estimation using RGB-D data. This system integrates object detection, segmentation, and pose estimation into a single …
-
New 3D Change Detection System for Retail Shelf Monitoring Introduced
Researchers have introduced ShelfChange3D, a new dataset and framework called ChangeBox for detecting changes in retail shelf inventory. This system operates in 3D space, identifying altered products and localizing them…
-
New agentic frameworks reconstruct 3D indoor scenes for simulation
Two new research papers introduce agentic frameworks for reconstructing 3D indoor scenes, aiming to create realistic and simulation-ready digital twins. LiteReality-Agent formulates 3D reconstruction as a coding problem…
-
New 'Finder' primitive enhances embodied AI object location
Researchers have introduced Finder, a new primitive designed for embodied agents to locate objects within 3D environments using language commands. Unlike previous methods that either couple search with exploration or re…
-
CALIPER framework uses RGB-D data for industrial part recognition
Researchers have developed CALIPER, a novel framework for recognizing visually similar industrial parts without requiring large, class-specific datasets or CAD models. This model-free approach utilizes RGB-D data to com…
-
Robotics research introduces outcome-based representation learning
Researchers have developed a novel method for learning manipulation-sufficient representations in robotics, focusing on action outcomes rather than dense geometric states. This approach utilizes an action-conditioned ou…
-
New E-RGB-D System Achieves Real-Time Color and Depth Perception at High Speeds
Researchers have developed E-RGB-D, a novel system that combines event-based cameras with a digital light processing projector to achieve real-time, high-speed RGB-D perception. This integration addresses the limitation…