Comments (7)
FeatDepth/mono/datasets/kitti_dataset.py
Line 19 in db7344a
Did you modify the normalized intrinsics here
from featdepth.
mono/datasets/folder_dataset.py
BTW, you can refer to this code to build your own dataloader for your custom dataset.
And you can show me your codes about dataloader, I will help you review it.
from featdepth.
Thank you for your kind response! My custom dataset is "Nuscenes" and there is camera intrinsic parameter for every sample, so I can call a function get_intrininsics in the getitem of the dataloader. As you said, I have build a custom dataloader for Nuscenes, very similar to your cityscape_dataset.py. However, I don't think that the problems lies in the dataloader, since this would not explain why I can get good results for some input sizes.
from featdepth.
Hi @cocacola0 .
I also met a similar problem. The input resolution for posecnn is hard coded here.
FeatDepth/mono/model/mono_fm_joint/net.py
Line 148 in db7344a
I think that's why you can get reasonable results when you keep the aspect ratio, but black with other sizes.
from featdepth.
Thank you @rxqy . I have already removed the hard coded resolution from there, still I am facing the same issue.
from featdepth.
Someone tell me that the issue can be solved in this way, 'Later I figured it out that there was nothing wrong with the code. Due to a change in nvidia-driver version, I was facing this issue. I was able to solve this issue by installing numba==0.43.1. ’
You may try.
from featdepth.
I will close this issue, you can reopen it if you have further question.
from featdepth.
Related Issues (20)
- running infer_singleimage.py in win10 ModuleNotFoundError: No module named 'resource' HOT 1
- it seems that the decoder of the auto-encoder network is not trained in the code HOT 1
- About the training and testing setting. HOT 4
- Is there a way to visually check the output of the FeatureNet HOT 1
- Question about MS evaluation HOT 5
- dataset.flag HOT 1
- An error occurred while training my own dataset HOT 2
- How to start non-distributed training HOT 2
- Why use DistOptimizerHook? HOT 1
- Problems using DDP HOT 2
- Train only Image Reconstruction Model HOT 2
- cfg_kitti_fm.py stops while training, expected 4 input channels, but got 3 channels instead
- train on own dataset with fm_joint.cfg HOT 1
- pose question HOT 1
- pose question HOT 2
- feature-metric loss only use the first output of the Autoencoder HOT 2
- Evaluation Issue HOT 1
- how to use multi-gpus training?
- Weights for Monocular-only training
- 关于online refinement的疑问
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from featdepth.