Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,54 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
datasets:
|
| 4 |
+
- LEI-QI-233/MicroG-4M
|
| 5 |
+
metrics:
|
| 6 |
+
- mAP
|
| 7 |
+
- F1-score
|
| 8 |
+
- Recall
|
| 9 |
+
- AUROC
|
| 10 |
+
pipeline_tag: video-classification
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Here stores all fine-tuned weights of our dataset.
|
| 14 |
+
|
| 15 |
+
## Please view our GitHub and dataset firstly:
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
<div align="left">
|
| 19 |
+
<a href="https://github.com/LEI-QI-233/HAR-in-Space"
|
| 20 |
+
style="display: inline-block; margin: 0 4px;">
|
| 21 |
+
<img src="https://img.shields.io/badge/GitHub-GitHub Repo-white?logo=github"
|
| 22 |
+
alt="GitHub"/>
|
| 23 |
+
</a>
|
| 24 |
+
<a href="https://huggingface.co/datasets/LEI-QI-233/MicroG-4M"
|
| 25 |
+
style="display: inline-block; margin: 0 4px;">
|
| 26 |
+
<img src="https://img.shields.io/badge/Hugging%20Face-Dataset-orange?logo=huggingface"
|
| 27 |
+
alt="Hugging Face Dataset"/>
|
| 28 |
+
</a>
|
| 29 |
+
</div>
|
| 30 |
+
|
| 31 |
+
---
|
| 32 |
+
|
| 33 |
+
### Performance comparison of models fine-tuned on MicroG-4M for HAR
|
| 34 |
+
|
| 35 |
+
| Arch | TC | Backbone | #Params (M) | mAP (%) | F1-score (%) | Recall (%) | AUROC (%) |
|
| 36 |
+
| -------- | ---- | -------- | ----------- | ------- | ------------ | ---------- | --------- |
|
| 37 |
+
| C2D | 8×8 | R50 | 23.61 | 29.51 | 8.09 | 6.58 | 83.49 |
|
| 38 |
+
| C2D NLN | 8×8 | R50 | 30.97 | 44.64 | 28.30 | 24.86 | 89.40 |
|
| 39 |
+
| I3D | 8×8 | R50 | 27.33 | 46.41 | 26.37 | 22.25 | 88.79 |
|
| 40 |
+
| I3D NLN | 8×8 | R50 | 34.68 | 47.12 | 28.07 | 24.65 | 88.52 |
|
| 41 |
+
| Slow | 8×8 | R50 | 31.74 | 45.19 | 26.13 | 22.77 | 88.49 |
|
| 42 |
+
| Slow | 4×16 | R50 | 31.74 | 46.37 | 28.72 | 25.38 | 88.30 |
|
| 43 |
+
| SlowFast | 8×8 | R50 | 33.76 | 43.02 | 22.63 | 18.98 | 88.51 |
|
| 44 |
+
| SlowFast | 4×16 | R50 | 33.76 | 42.10 | 23.69 | 20.18 | 87.54 |
|
| 45 |
+
| MViTv1 | 16×4 | B-CONV | 36.34 | 12.86 | 5.54 | 4.66 | 74.63 |
|
| 46 |
+
| MViTv2 | 16×4 | S | 34.27 | 15.14 | 8.16 | 7.17 | 78.61 |
|
| 47 |
+
| X3D | 13×6 | S | 2.02 | 14.07 | 5.77 | 4.52 | 78.23 |
|
| 48 |
+
| X3D | 16×5 | L | 4.37 | 18.70 | 9.15 | 7.47 | 78.27 |
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
**Note:**
|
| 52 |
+
- All models has been pretrained on Kinetics400 dataset and continually trained on MicroG-4M.
|
| 53 |
+
- `TC` denotes the temporal configuration (frame length × sampling rate).
|
| 54 |
+
- `#Params` indicates the number of parameters (in millions, M).
|