Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
machineuser
commited on
Commit
·
4e75983
1
Parent(s):
7b4cf0b
Sync widgets demo
Browse files
packages/tasks/src/tasks/mask-generation/about.md
CHANGED
@@ -6,7 +6,7 @@ When filtering for an image, the generated masks might serve as an initial filte
|
|
6 |
|
7 |
### Masked Image Modelling
|
8 |
|
9 |
-
Generating masks can be done to facilitate learning, especially in semi- or unsupervised learning. For example, the [BEiT](https://huggingface.co/docs/transformers/model_doc/beit)
|
10 |
|
11 |
### Human-in-the-loop
|
12 |
|
@@ -25,7 +25,6 @@ from transformers import pipeline
|
|
25 |
generator = pipeline("mask-generation", device = 0, points_per_batch = 256)
|
26 |
image_url = "https://huggingface.co/ybelkada/segment-anything/resolve/main/assets/car.png"
|
27 |
outputs = generator(image_url, points_per_batch = 256)
|
28 |
-
|
29 |
```
|
30 |
|
31 |
## Useful Resources
|
|
|
6 |
|
7 |
### Masked Image Modelling
|
8 |
|
9 |
+
Generating masks can be done to facilitate learning, especially in semi- or unsupervised learning. For example, the [BEiT model](https://huggingface.co/docs/transformers/model_doc/beit) uses image-masked patches in the pre-training.
|
10 |
|
11 |
### Human-in-the-loop
|
12 |
|
|
|
25 |
generator = pipeline("mask-generation", device = 0, points_per_batch = 256)
|
26 |
image_url = "https://huggingface.co/ybelkada/segment-anything/resolve/main/assets/car.png"
|
27 |
outputs = generator(image_url, points_per_batch = 256)
|
|
|
28 |
```
|
29 |
|
30 |
## Useful Resources
|