Comments (3)
houqb
commented on July 18, 2024
We process each image in a sliding window way. As mentioned in the paper, we use the UperNet head as our decoder head.
from volo.
ajtao
commented on July 18, 2024
right thanks.
from volo.
ajtao
commented on July 18, 2024
@Andrew-Qibin I'm seeing rather poor Cityscapes segmentation results (training starts at 20 IOU first epoch and only gets to 53 IOU) right out of the box using a volo_d2 trunk (using imagenet pretrained weight). Probably i've got some tensor ordering wrong or something, but is there any trick to adapting the code to higher resolution? I had to of course override the positional encodings from the checkpoint with a new higher resolution positional encoding (1024). And i created a new forward() that supplies the features in N, C, H, W form. But hmm, not sure what i've got wrong right now.
from volo.
Related Issues (20)
- Halonets declare a SOTA in CVPR, any contribution to conference?
- Some thoughts about volo
- Colab Notebook doesn't work & gives wrong results. HOT 4
- Finetune with 512 image size HOT 2
- Balanced class weight HOT 1
- Will you release the codes for Semantic Segmentation?
- AttributeError: 'tuple' object has no attribute 'log_softmax' HOT 2
- Could you share the training script for training volo_d5 ↑512?
- run_error HOT 1
- SystemError: returned NULL without setting an error
- May you provide the FLOPs calculation script ?
- size mismatch for pos_embed: copying a param with shape torch.Size([1, 14, 14, 768]) from checkpoint, the shape in current model is torch.Size([1, 14, 14, 512]). HOT 1
- RuntimeError: The size of tensor a (28) must match the size of tensor b (14) at non-singleton dimension 2 HOT 2
- is the order of v_pj(x) result in algoritm1 correct ?
- The code for Semantic Segmentation?
- When training own dataset, an error occurs when changing numberclasses to the corresponding category. If it is the default, it will report an error
- volo-d1 training without token label data HOT 1
- pre trained model file is broken. HOT 1
- Question about computational complexity formulation of Outlooker Attention
- Ablation study and official code problem
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 volo.