Towards comprehensive multi-task land cover change detection leveraging vision-language model and LLM-driven agents
Multi-task land-cover change detection with vision-language models and LLM-driven agents for urban remote sensing.
tengxi-wang
•
•
1 min read
DSTI-Net: A Dynamic Spatial–Temporal Interaction Network With Semantic Guidance for 2-D and 3-D Change Detection
This paper proposes DSTI-Net, a multitask framework integrating SAM-based semantic guidance and dynamic spatial-temporal interaction for joint 2-D and 3-D change detection from …
tengxi-wang
Point and voxel cross perception with lightweight cosformer for large-scale point cloud semantic segmentation
JAG 2024 · First author. PVCFormer combines point and voxel representations with lightweight cross-attention for large-scale outdoor semantic segmentation, alongside the SYSU9 …
shuai-zhang
Chat3D: Interactive understanding 3D scene-level point clouds by chatting with foundation model for urban ecological construction
This paper presents Chat3D, a method for applying large language models to urban ecological construction by combining the results of 3D point cloud semantic segmentation. The …
yiping-chen
SPTNet: Sparse Convolution and Transformer Network for Woody and Foliage Components Separation From Point Clouds
This paper proposes SPTNet, a novel deep learning architecture combining sparse convolution and transformer blocks for automated separation of woody and foliage components from …
shuai-zhang
3D Detection and Morphological Analysis of Extraterrestrial Lava Tubes: A Case Study in Haikou, China
This study presents a 3D detection workflow using SLAM100 handheld LiDAR for lava tube exploration, including point cloud processing, 3D modeling, and skylight detection, providing …
du-yumeng
SLAM-Based Forest Plot Mapping by Integrating IMU and Self-Calibrated Dual 3-D Laser Scanners
This paper proposes a dual-scanner backpack LiDAR system with automatic self-calibration and a tightly-coupled SLAM framework integrating IMU and forest natural features, achieving …
dong-pan
Learning point cloud context information based on 3D transformer for more accurate and efficient classification
This paper proposes a novel 3D transformer architecture for point cloud classification that captures local context information by highlighting feature importance differences …
yiping-chen