Closed ZrrSkywalker closed 4 months ago
Thank you very much for your attention. PointBind is an excellent project. We believe it has made a significant contribution to aligning point cloud and text modalities. The reason we didn't initially provide a citation is that PointBind was originally affiliated with llama-adapter (ImageBing-LLM). Later, it transitioned to OpenGVlab, and we couldn't locate a suitable citation path at that time.
We will soon update the GitHub page, further improving the code and resolving citation issues. We hope to contribute to the point cloud community together with you.
We sincerely appreciate your understanding :). Hope to make the 3D community better together with you!
Hi, thanks for this great project PointBLIP that directly aligns 3D with LLM!
Here is our project Point-Bind that first aligns 3D point clouds with LLM and other modalities, which is released in June, 2023. We also conduct 3D question answering, 3D retrieval and text-3D generation tasks. Given the high similarity of our two projects concerning idea (3D LLM) and implementation (a projection layer), we would be highly appreciated that you can consider citing our project. We will also cite yours.
The characters of Point-Bind are as follows:
The following are our examples of 3D question answering:
Thanks very much for your understanding!