TangYuan96

Hi, thank you for sharing such excellent results I would like to ask what variant do we use in few-shot learning on ScanObjectNN? OBJ ONLY OBJ BG PB T25 PB...

The Q-Former weights of X-InstructBLIP can not be download

5

the link `https://storage.googleapis.com/sfr-xinstructblip-data-research/model/xinstructblip_checkpoints/vicuna7b/image_qformer.pth` can not be accessed. Looking forward to your reply！！

would you share the dataset of pretraining for 3D object detection ?

1

First, thanks you for sharing your outstanding work ! would you share the dataset of pretraining for 3D object detection ? The whole Scannet dataset is too big to download...

How to reproducing the zero-shot result 49.38% on ModelNet40 ?

3

First of all, thanks for sharing your outstanding work I try to train zero-shot (python pretraining. py) on ModelNet40Align and ModelNet40Ply, the results can only reach 36.71% and 32.70% But...

Hello, dear friend! First, thanks for sharing so outstanding work ! ! I notice that Tabel 8 of paper report the recognition accuracy (%) **averaged** under the instruction-typed prompt “What...

Thank you for sharing so great work!! It is very powerful. Regarding the checkpoints for ShapeLLM, there seem to be two types of [checkpoints](https://github.com/qizekun/ShapeLLM/blob/main/docs/MODEL_ZOO.md) used in the evaluation scripts When...