Comments (1)
Hi, @chqzeng
Sounds like your data are similar to SemanticKITTI. We don't provide a specific script for custom data, but you can refer to our codes used in SemanticKITTI. You also need to construct superpoints similar to the codes under "data_prepare."
Our unsupervised method relies on deep networks, so it requires training. We get the pseudo labels with the help of K-Means, then train the deep networks with their supervision to enhance the discrimination of the network's features. Consequently, the features will be increasingly discriminative and clearly better than a random initialized network. So, running without training doesn't work. And I recommend reading our paper at first, as it will facilitate a better understanding of the methodology.
The output should be a folder containing the checkpoints as well as a log file. This log file records training loss, the k-means error, the accuracy of superpoints and primitives, and segmentation results on the validation set. You can find them in our paper.
from growsp.
Related Issues (20)
- Training ERROR After 280 epoches HOT 4
- How many GPUs are needed to complete this experiment, and how much video memory is needed HOT 4
- train_S3DIS.py error HOT 8
- How to Apply a Trained S3DIS Model to a Different Dataset for Inference or Evaluation? HOT 1
- Can you share the visual code
- Issue with mimicking S3DIS data using custom point cloud data HOT 1
- train my datasets HOT 2
- Superpoint acquisition HOT 1
- Cuda memory error HOT 2
- how to compute cost matrix in Hungarian algorithm? HOT 2
- An error occurs when calculating the loss function, suggesting that the batch_size is inconsistent at around 20+ epochs, what's going on?
- cudaErrorIllegalAddress: an illegal memory access was encountered
- visualization HOT 2
- Time cost HOT 2
- About the training. HOT 1
- Evaluation procedure HOT 4
- About pip install -U MinkowskiEngine --install-option="--blas=openblas" -v --no-deps HOT 1
- About Custom DataSet HOT 3
- CUDA out of memory HOT 3
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 growsp.