Comments (1)
Hi,
Thanks for your message and reproduction! I think there are three reasons that may lead to this discrepancy:
- general training stochasticities, like different pytorch/cuda versions or OS or different GPU architecture. Our papers' results were produced using a single NVIDIA V100 32GB GPU without distributed training.
- checkpoint selection rule. We were using the performance on the validation set to select the best checkpoint during training. I believe a better model could be found if we directly evaluated its performance on the testing set.
- moreover, the PnP solver may include another level of randomness. We borrow the improved PnP solver from the dsacstar as it is. Previously, we also found that the PnP solver could give different results on different machines (with the same predicted scene coordinates).
Lastly, we provided our model checkpoints at https://1drv.ms/u/s!AnkbqTET-eNqgoRsgBXkEg-PFSqudA?e=S6Pf43. You can also give it a try to rule out some possibilities.
from crossloc.
Related Issues (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 crossloc.