Posts

You can build your own LLM with simple setup. Without prioer knowledge, you can build your local LLM. Here is procedure in Windows 1 Install Ollama 1.1 download and install Ollama 3.1 https://ollama.com/ 1.2 Choose the model 1.3 There are 8b, 70b, 405b options. for self use, 8b is enough. 1.4 copy the comand to powersell to install the model 2 Insall openweb UI 2.1 install Docker https://www.docker.com/ (If can not install, try to install previous version)...

PointNetLK: Lessons Learned for Beginners in Deep Learning and Point Cloud Registration PointNetLK is built upon the PointNet model, a novel and widely-used approach for extracting features from point clouds, originally developed by Zhang et al. PointNetLK applies the Lucas-Kanade (LK) algorithm to approximate the transformation matrix between point clouds. For beginners in deep learning or point cloud registration, here are some important lessons: Clarify Your Use Case It’s crucial to understand your target application and how users will interact with your model....

The X ArUco marker is based on the ArUco marker but includes two additional pattern types. The four circles are used to create auxiliary diagonals for line fitting, while the four squares are employed for sub-pixel corner detection to achieve higher precision. Inspired by An improved ArUco marker for monocular vision ranging and ArUcoE: Enhanced ArUco Marker Combine the line fitting and sub-pixel methods. Re-projection error imporved almost 70%...