Kitti
PulseAugur coverage of Kitti — every cluster mentioning Kitti across labs, papers, and developer communities, ranked by signal.
13 day(s) with sentiment data
-
New ADM-Fusion method enhances ego-motion estimation with adaptive multi-sensor fusion
Researchers have developed ADM-Fusion, a novel deep learning approach for ego-motion estimation that adaptively fuses data from multiple sensors. This method utilizes a mixture-of-experts framework with content-aware ro…
-
Pocket-SLAM tackles memory limits in 3DGS-SLAM for autonomous driving
Researchers have developed Pocket-SLAM, a novel method to improve the memory efficiency of 3D Gaussian Splatting for Simultaneous Localization and Mapping (SLAM). This approach addresses the issue of accumulating Gaussi…
-
Polycepta framework enhances multi-object tracking with dynamic appearance estimation
Researchers have developed Polycepta, a novel framework for object-centric appearance state estimation in multi-object tracking. Unlike traditional methods that use static descriptors, Polycepta recursively estimates an…
-
New CABLE framework boosts LMM efficiency for V2X systems
Researchers have developed CABLE, a novel framework designed to enhance the efficiency of large multimodal models (LMMs) in vehicle-to-everything (V2X) systems. This system reduces communication overhead and cloud-side …
-
New RAMS system adapts YOLOv8 tiers for edge AI perception
Researchers have developed RAMS, a novel runtime controller designed for embedded edge perception systems. RAMS dynamically switches between different tiers of YOLOv8 models based on real-time device resource monitoring…
-
MVOFormer Transformer Boosts Monocular Visual Odometry Robustness
Researchers have introduced MVOFormer, a new transformer-based framework designed to enhance monocular visual odometry (MVO) for autonomous navigation. This model integrates geometric motion cues with semantic object pr…
-
DiskChunGS enables large-scale 3D Gaussian SLAM via disk memory management
Researchers have developed DiskChunGS, a novel 3D Gaussian Splatting SLAM system designed to overcome GPU memory limitations for large-scale 3D reconstructions. By employing an out-of-core approach, the system stores in…
-
Researchers analyze training-free methods for object detection dataset errors
Researchers have analyzed the effectiveness of training-free methods for detecting annotation errors in object detection datasets. Their findings indicate that these methods are adept at identifying semantic mislabeling…
-
Satellite imagery resolves scale ambiguity in 3D reconstruction models
Researchers have developed a new method to resolve the scale ambiguity in feed-forward 3D reconstruction models by utilizing satellite imagery as a global metric reference. This approach integrates satellite patches wit…
-
New method matches 2D polygons for pose estimation
Researchers have introduced a novel Zero-shot Polygon Matching paradigm with Pre-trained Models (Z(PM)2) to address the challenges of matching 2D polygons in stereo imagery. This method leverages pre-trained models like…
-
LiAuto-GeoX transformer enables real-time 3D driving scene understanding
Researchers have developed LiAuto-GeoX, an efficient transformer model for real-time 3D scene understanding in autonomous driving. This model leverages sparse LiDAR data for geometric grounding and employs a novel disti…
-
CIPER framework unifies image retrieval and pose estimation
Researchers have developed CIPER, a novel unified framework for cross-view geo-localization that simultaneously performs large-scale image retrieval and precise pose estimation. Unlike previous methods that handled thes…
-
CIPER framework unifies image retrieval and pose estimation
Researchers have introduced CIPER, a novel framework designed to unify cross-view geo-localization tasks. This system simultaneously performs city-scale image retrieval and precise 3-degree-of-freedom pose estimation by…
-
PillarDETR advances real-time 3D object detection for autonomous driving
Researchers have introduced PillarDETR, a new architecture for real-time 3D object detection, particularly for autonomous driving systems. This model integrates a YOLOv8-derived backbone with an RT-DETR decoder, optimiz…
-
New LFA method enhances AI error prediction for self-driving cars
Researchers have developed a new method called Layer Feature Attention (LFA) to improve the introspection of 2D object detectors used in automated driving systems. LFA utilizes an attention mechanism to aggregate featur…
-
New method achieves joint multi-camera LiDAR calibration
Researchers have developed a novel two-stage framework for joint multi-camera LiDAR extrinsic calibration. This method first uses CMRNext for initial pairwise estimates and then refines them using a multi-frame bundle a…
-
New Transformers Enhance 3D Scene Reconstruction and Edge Deployment
Researchers have developed new transformer-based models for 3D scene reconstruction from visual inputs. DVGT, a Driving Visual Geometry Transformer, reconstructs dense 3D point maps from unposed multi-view images withou…
-
SA4Depth improves self-supervised monocular depth estimation
Researchers have introduced SA4Depth, a novel approach to enhance self-supervised monocular depth estimation. This method focuses on improving the alignment between the scale estimates from separate depth and pose netwo…
-
New SLIM Model Enhances Long-Range Driving Depth Estimation with Sparse LiDAR
Researchers have developed SLIM (Sparse-LiDAR Injected Monocular geometry), a novel approach to enhance monocular depth estimation for long-range driving scenarios. SLIM adapts the MoGe-2 model to directly incorporate s…
-
New method refines LiDAR-camera calibration using evidence maps
Researchers have developed a new method for calibrating LiDAR and camera systems, particularly for agricultural environments. This approach uses a "support-map-driven" technique to identify which observations are most c…