Update README.md
Browse files
README.md
CHANGED
|
@@ -18,7 +18,11 @@ This repository contains the model weights for **Concerto**, a novel approach fo
|
|
| 18 |
- **Inference:** [https://github.com/Pointcept/Concerto](https://github.com/Pointcept/Concerto)
|
| 19 |
|
| 20 |
## Models
|
| 21 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
| Model Size | Channels | Depths | Head nums |
|
| 23 |
| --- | --- | --- | --- |
|
| 24 |
| Tiny | (16, 32, 64, 128, 256) | (1, 1, 1, 3, 1) | (1, 2, 4, 8, 16) |
|
|
|
|
| 18 |
- **Inference:** [https://github.com/Pointcept/Concerto](https://github.com/Pointcept/Concerto)
|
| 19 |
|
| 20 |
## Models
|
| 21 |
+
concerto_base_siglip.pth: replace DINOv2 with SigLIPv2 in cross-modal distillation.
|
| 22 |
+
concerto_large_linear_prob_head_sc.pth: ScanNet linear probing head.
|
| 23 |
+
concerto_large_outdoor.pth: outdoor pretraining weight using similar recipe.
|
| 24 |
+
|
| 25 |
+
The default models(concerto_large/base/small/tiny) are the pre-release version of our next work, Utonia, which can deal with input without color and normal. If colors and normals are not available, please set them to zeros. We pre-release these for general public use because many tasks lack such information. The original Concerto model is `concerto_base_origin.pth`.
|
| 26 |
| Model Size | Channels | Depths | Head nums |
|
| 27 |
| --- | --- | --- | --- |
|
| 28 |
| Tiny | (16, 32, 64, 128, 256) | (1, 1, 1, 3, 1) | (1, 2, 4, 8, 16) |
|