topo-epfl / crossloc Goto Github PK
View Code? Open in Web Editor NEW[CVPR'22] CrossLoc localization: a cross-modal visual representation learning method for absolute localization
License: MIT License
[CVPR'22] CrossLoc localization: a cross-modal visual representation learning method for absolute localization
License: MIT License
Hi,
Thanks for the code, I'm very interested in your outstanding work!!!
I have tried to retrain your network, and the result shows that the index values of in-place localization accuracy I reproduced over urbanscape dataset are better than those given in your paper. So I want to ask you what might cause this situation.
Looking forward to your reply.
Here is the result over urbanscape dataset from your paper:
Here is quantitative comparison between paper's result and my reproduced result:
Hi scholars, hello everyone. When I am executing -- task semantics/depth/normal, regardless of whether I change the weight name you provided to mlr_encoder_ * or ckpt_iter_ *, the following error still appears, indicating that the information was not read correctly. How should this situation be resolved? If I execute -- task coord, this problem won't occur. Why is this?
"(crossloc) user@user:~/workspaces/CrossLoc$ python3 test_single_task.py naturescape --task semantics --uncertainty mle --section test_drone_real --network_in /media/user/Data/CrossLoc-Datasets/ckpt_CrossLoc_network_weights/naturescape/in-place/
Namespace(fullsize=False, grayscale=False, hypotheses=64, inlieralpha=100, keywords=None, max_ckpt_iter=None, maxpixelerror=100, min_ckpt_iter=None, network_in='/media/slby/user/CrossLoc-Datasets/ckpt_CrossLoc_network_weights/naturescape/in-place/', plot=False, save_pred=False, scene='naturescape', search_dir=False, section=['test_drone_real'], session='', task='semantics', threshold=10, tiny=False, uncertainty='MLE')
To load network weight from /media/user/Data/CrossLoc-Datasets/ckpt_CrossLoc_network_weights/naturescape/in-place
With the keywords None, and min ckpt iter None, max ckpt iter None the following 1 network weight paths are retrieved:
Network weight #0: /media/user/Data/CrossLoc-Datasets/ckpt_CrossLoc_network_weights/naturescape/in-place/ckpt_iter_crossloc_se_naturescape_in-place.net
==================== Evaluating network #1 / 1 ====================
Path: /media/user/Data/CrossLoc-Datasets/ckpt_CrossLoc_network_weights/naturescape/in-place/ckpt_iter_crossloc_se_naturescape_in-place.net
Traceback (most recent call last):
File "test_single_task.py", line 420, in
main()
File "test_single_task.py", line 299, in main
network = config_network(scene, task, tiny, grayscale, uncertainty, fullsize, network_path, num_enc=num_enc)
File "/home/user/workspaces/CrossLoc/utils/evaluation.py", line 162, in config_network
raise NotImplementedError
NotImplementedError"
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.