Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: en
|
| 3 |
+
license: mit
|
| 4 |
+
tags:
|
| 5 |
+
- atc
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# Fine-Tuned Agglomerative Token Clustering - DeiT-Small-Single - ImageNet-1k
|
| 9 |
+
|
| 10 |
+
### Model Details
|
| 11 |
+
Agglomerative Token Clustering (ATC), a novel hierarchical hard-merging based token reduction method.
|
| 12 |
+
|
| 13 |
+
- **Developed by:** Joakim Bruslund Haurum, Sergio Escalera, Graham W. Taylor, and Thomas B. Moeslund
|
| 14 |
+
- **Model type:** Vision Transformer
|
| 15 |
+
- **License:** MIT
|
| 16 |
+
- **Task:** Image Classification
|
| 17 |
+
|
| 18 |
+
### Model Card
|
| 19 |
+
- **Backbone:** DeiT-Small
|
| 20 |
+
- **Token Reduction Method:** ATC
|
| 21 |
+
- **Linkage Function:** Single
|
| 22 |
+
- **Reduction Ratio:** {0.25, 0.5, 0.7, 0.9}
|
| 23 |
+
- **Reduction Stages:** 3, 6, 9
|
| 24 |
+
|
| 25 |
+
### More Resources
|
| 26 |
+
- **Repository:** [https://github.com/JoakimHaurum/ATC](https://github.com/JoakimHaurum/ATC)
|
| 27 |
+
- **Paper:** [TBD](TBD)
|
| 28 |
+
- **Project Page:** [https://vap.aau.dk/atc](https://vap.aau.dk/atc)
|
| 29 |
+
- **HuggingFace Collection:** [https://huggingface.co/collections/joakimbh/agglomerative-token-clustering-66e94dfb313e85ec97590fe4](https://huggingface.co/collections/joakimbh/agglomerative-token-clustering-66e94dfb313e85ec97590fe4)
|
| 30 |
+
|
| 31 |
+
### Use
|
| 32 |
+
The model files contain both standard and EMA model parameters. The version which gave the best performance is indicated with the "ema_best" boolean.
|