InstantMesh

InstantMesh copied to clipboard

Training on NeRF representation only returns white images for the rendered images and no depth maps.

1

I have tried to train on the NeRF representation, but the generator only returns white rendered images.   Can someone please help me? Are the weights of the lrm_reconstructor...

SBlumenstock1

Custom Dataset preparing for training

2

Your model's ability to efficiently render high-quality 3D models from a single image is highly impressive and practically significant. We are very interested in this work, and we would like...

ustbzgn

Loading diffusion model ... Loading pipeline components...: 12%

1

**/tmp/tmppngxpwds.obj Traceback (most recent call last):** File "/home/jkx/anaconda3/envs/InstantMesh/lib/python3.10/site-packages/gradio/queueing.py", line 536, in process_events response = await route_utils.call_process_api( File "/home/jkx/anaconda3/envs/InstantMesh/lib/python3.10/site-packages/gradio/route_utils.py", line 276, in call_process_api output = await app.get_blocks().process_api( File "/home/jkx/anaconda3/envs/InstantMesh/lib/python3.10/site-packages/gradio/blocks.py", line 1923,...

nagexiaochengzi

Which model to use for image normals generation?

3

There is a script in zero123++ repository for image normal generation but it is used as a controller with zero123++ which generate 6 view images. Is there any other model...

asadullahnaeem

Thank you for your great work! What if I only have one image and I want to train a mesh that renders the same view as my original image? Need...

Dumeowmeow

Run Run. py to generate 6 different angle images with empty output images and pure black images. Guess it was a sampling error? There is no problem printing the input images. Could you please help me take a look, experts? Thank you

1

RuntimeWarning:invalid value encountered in cast images = (images * 255).round().astype("unint8")

anshanYCL

IndexError: too many indices for tensor of dimension 2

1

你好，我在运行训练代码，model.py文件里面这一段： input_c2ws=torch.tensor(batch['input_c2ws']) print(input_c2ws.shape) #nput_c2ws=np.array(batch['input_c2ws'].cpu()) input_c2ws=input_c2ws.reshape(input_c2ws.shape[0],-1) #input_c2ws=input_c2ws.flatten(-2) input_Ks = batch['input_Ks'].flatten(-2) input_extrinsics = input_c2ws[:, :, :12]， 报了IndexError: too many indices for tensor of dimension 2的错误，我的input_c2ws是6*4*4的list，所以源代码的flatten操作我改成了reshape，变换后应该是6*16，但是input_c2ws[:, :, :12]这一行就执行不了，所以想问问看您之前跑的时候的input_c2ws的维度具体是多少呢？是我这边的维度有错么

throb081

Number of epochs before seeing a decrease in the loss

7

Hi, I am currently training the instant-mesh model but after 3 epochs I am not seeing convergence and the loss does not seem to decrease. I am using the same...

Kev1MSL

Train on nerf of stage 1 with custom dataset

10

Hi, guys, i rendered my own rings dataset with 32 views , for every view, i have a RGBA image, where A means mask, normal image and depth image, and...

hopeliu20160622

Train zero123 with own dataset

2

I wonder if you can provide any code snippet on how to create a custom dataset for fine-tuning. If I understood the fine-tuning code right the model expects the original...

Maxbr2021