search

autonomousvision / occupancy_networks

This repository contains the code for the paper "Occupancy Networks - Learning 3D Reconstruction in Function Space"
https://avg.is.tuebingen.mpg.de/publications/occupancy-networks
MIT License
1.51k stars 292 forks source link

Use my images to test, and get bad visual #74

Open fashionguy opened 4 years ago

fashionguy commented 4 years ago

I tested ONet with my own images and got bad results. What should I do to be as perfect as the demo you provided. my images as below. 09 10 11

fashionguy commented 4 years ago

results respectively as below. image image image

fashionguy commented 4 years ago

meshlab

------------------ 原始邮件 ------------------ 发件人: "autonomousvision/occupancy_networks" <notifications@github.com>; 发送时间: 2020年9月21日(星期一) 凌晨3:44 收件人: "autonomousvision/occupancy_networks"<occupancy_networks@noreply.github.com>; 抄送: "Stan"<3301225509@qq.com>;"Author"<author@noreply.github.com>; 主题: Re: [autonomousvision/occupancy_networks] Use my images to test, and get bad visual (#74)

how do you visualize the generated .off files

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

georged4s commented 3 years ago

yea agree. some custom images even from its training sets doesn't perform well neither. Not sure if I have missed anything? image image

gokceay commented 1 year ago

I tested ONet with my own images and got bad results. What should I do to be as perfect as the demo you provided. my images as below. 09 10 11

Hi @fashionguy I want to test my custom dataset and I created my 3D files .binvox, pointcloud.npz and point.npz, but I am not sure about images of the 3D data. What is the format of the images of the training for single image reconstruction (resnet18)? npz or png? For each 3D off watertight mesh I have one single real png image. There is a get_r2n2_cameras.py but I could not clearly understand what is happening inside. Any help would be greatly appreciated. Thanks in advance

jndmw-111 commented 1 year ago

Hello! I have the same question, is it clear to you now what are the restrictions on the input image

TaroPlay commented 1 year ago

I tested ONet with my own images and got bad results. What should I do to be as perfect as the demo you provided. my images as below. 09 10 11

Hi @fashionguy I want to test my custom dataset and I created my 3D files .binvox, pointcloud.npz and point.npz, but I am not sure about images of the 3D data. What is the format of the images of the training for single image reconstruction (resnet18)? npz or png? For each 3D off watertight mesh I have one single real png image. There is a get_r2n2_cameras.py but I could not clearly understand what is happening inside. Any help would be greatly appreciated. Thanks in advance

halo,my friend,I knew you created the pointcloud.npz and point.npz about your dataset.I faced the difficulty in creating the pointcloud.npz and point.npz files.I'd really appreciate it if you could lend me a hand.