Spaces:
Runtime error
Runtime error
Commit
•
1b20dc8
1
Parent(s):
b917128
Delete clipseg/Readme.md
Browse files- clipseg/Readme.md +0 -84
clipseg/Readme.md
DELETED
@@ -1,84 +0,0 @@
|
|
1 |
-
# Image Segmentation Using Text and Image Prompts
|
2 |
-
This repository contains the code used in the paper ["Image Segmentation Using Text and Image Prompts"](https://arxiv.org/abs/2112.10003).
|
3 |
-
|
4 |
-
**The Paper has been accepted to CVPR 2022!**
|
5 |
-
|
6 |
-
<img src="overview.png" alt="drawing" height="200em"/>
|
7 |
-
|
8 |
-
The systems allows to create segmentation models without training based on:
|
9 |
-
- An arbitrary text query
|
10 |
-
- Or an image with a mask highlighting stuff or an object.
|
11 |
-
|
12 |
-
### Quick Start
|
13 |
-
|
14 |
-
In the `Quickstart.ipynb` notebook we provide the code for using a pre-trained CLIPSeg model. If you run the notebook locally, make sure you downloaded the `rd64-uni.pth` weights, either manually or via git lfs extension.
|
15 |
-
It can also be used interactively using [MyBinder](https://mybinder.org/v2/gh/timojl/clipseg/HEAD?labpath=Quickstart.ipynb)
|
16 |
-
(please note that the VM does not use a GPU, thus inference takes a few seconds).
|
17 |
-
|
18 |
-
|
19 |
-
### Dependencies
|
20 |
-
This code base depends on pytorch, torchvision and clip (`pip install git+https://github.com/openai/CLIP.git`).
|
21 |
-
Additional dependencies are hidden for double blind review.
|
22 |
-
|
23 |
-
|
24 |
-
### Datasets
|
25 |
-
|
26 |
-
* `PhraseCut` and `PhraseCutPlus`: Referring expression dataset
|
27 |
-
* `PFEPascalWrapper`: Wrapper class for PFENet's Pascal-5i implementation
|
28 |
-
* `PascalZeroShot`: Wrapper class for PascalZeroShot
|
29 |
-
* `COCOWrapper`: Wrapper class for COCO.
|
30 |
-
|
31 |
-
### Models
|
32 |
-
|
33 |
-
* `CLIPDensePredT`: CLIPSeg model with transformer-based decoder.
|
34 |
-
* `ViTDensePredT`: CLIPSeg model with transformer-based decoder.
|
35 |
-
|
36 |
-
### Third Party Dependencies
|
37 |
-
For some of the datasets third party dependencies are required. Run the following commands in the `third_party` folder.
|
38 |
-
```bash
|
39 |
-
git clone https://github.com/cvlab-yonsei/JoEm
|
40 |
-
git clone https://github.com/Jia-Research-Lab/PFENet.git
|
41 |
-
git clone https://github.com/ChenyunWu/PhraseCutDataset.git
|
42 |
-
git clone https://github.com/juhongm999/hsnet.git
|
43 |
-
```
|
44 |
-
|
45 |
-
### Weights
|
46 |
-
|
47 |
-
The MIT license does not apply to these weights.
|
48 |
-
|
49 |
-
We provide two model weights, for D=64 (4.1MB) and D=16 (1.1MB).
|
50 |
-
```
|
51 |
-
wget https://owncloud.gwdg.de/index.php/s/ioHbRzFx6th32hn/download -O weights.zip
|
52 |
-
unzip -d weights -j weights.zip
|
53 |
-
```
|
54 |
-
|
55 |
-
|
56 |
-
### Training and Evaluation
|
57 |
-
|
58 |
-
To train use the `training.py` script with experiment file and experiment id parameters. E.g. `python training.py phrasecut.yaml 0` will train the first phrasecut experiment which is defined by the `configuration` and first `individual_configurations` parameters. Model weights will be written in `logs/`.
|
59 |
-
|
60 |
-
For evaluation use `score.py`. E.g. `python score.py phrasecut.yaml 0 0` will train the first phrasecut experiment of `test_configuration` and the first configuration in `individual_configurations`.
|
61 |
-
|
62 |
-
|
63 |
-
### Usage of PFENet Wrappers
|
64 |
-
|
65 |
-
In order to use the dataset and model wrappers for PFENet, the PFENet repository needs to be cloned to the root folder.
|
66 |
-
`git clone https://github.com/Jia-Research-Lab/PFENet.git `
|
67 |
-
|
68 |
-
|
69 |
-
### License
|
70 |
-
|
71 |
-
The source code files in this repository (excluding model weights) are released under MIT license.
|
72 |
-
|
73 |
-
### Citation
|
74 |
-
```
|
75 |
-
@InProceedings{lueddecke22_cvpr,
|
76 |
-
author = {L\"uddecke, Timo and Ecker, Alexander},
|
77 |
-
title = {Image Segmentation Using Text and Image Prompts},
|
78 |
-
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
|
79 |
-
month = {June},
|
80 |
-
year = {2022},
|
81 |
-
pages = {7086-7096}
|
82 |
-
}
|
83 |
-
|
84 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|