describe the model
Browse files
README.md
CHANGED
|
@@ -3,4 +3,18 @@ license: gpl-3.0
|
|
| 3 |
pipeline_tag: image-classification
|
| 4 |
---
|
| 5 |
|
| 6 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
pipeline_tag: image-classification
|
| 4 |
---
|
| 5 |
|
| 6 |
+
K-Nearest-Neighbour model of human perception of Amsterdam street view imagery. data.npz is a compressed numpy file with 10 records, it can be loaded with numpy.load and the resulting object can be indexed by any of the following keys:
|
| 7 |
+
|
| 8 |
+
* walkability_vecs, walkability_scores
|
| 9 |
+
* bikeability_vecs, bikeability_scores
|
| 10 |
+
* pleasantness_vecs, pleasantness_scores
|
| 11 |
+
* greenness_vecs, greenness_scores
|
| 12 |
+
* safety_vecs, safety_scores
|
| 13 |
+
|
| 14 |
+
The _vecs entries are matrices of size Nx1024 and the _scores entries are vectors of size N.
|
| 15 |
+
|
| 16 |
+
The _vecs are encoded by OpenCLIP ViT-H-14-378-quickgelu (pretrained: dfn5b).
|
| 17 |
+
|
| 18 |
+
The _scores are the given rating scores (1 to 5) for the corresponding vector. For example, the vector given by walkability_vecs[10,:] has a corresponding score in walkability_scores[10].
|
| 19 |
+
|
| 20 |
+
This can be used to model a score for any given vector from an image encoded by the above CLIP model.
|