Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -119,7 +119,8 @@ with gr.Blocks() as demo:
|
|
| 119 |
|
| 120 |
<center>by <a href="https://www.tonyassi.com/" target="_blank">Tony Assi</a></center>
|
| 121 |
|
| 122 |
-
|
|
|
|
| 123 |
<center> This is a demo of product recommendation using image similarity of user preferences. </center>
|
| 124 |
|
| 125 |
The the user selects their favorite product which then gets added to the user preference group. Each of the image embeddings in the user preference products get averaged into a preference embedding. Each round some products are displayed: 5 products most similar to user preference embedding and 5 random products. Embeddings are generated with [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224). The dataset used is [tonyassi/finesse1-embeddings](https://huggingface.co/datasets/tonyassi/finesse1-embeddings).
|
|
|
|
| 119 |
|
| 120 |
<center>by <a href="https://www.tonyassi.com/" target="_blank">Tony Assi</a></center>
|
| 121 |
|
| 122 |
+
<center>by <a href="https://huggingface.co/blog/tonyassi/product-recommendation-using-image-similarity/" target="_blank">Read the Article</a></center>
|
| 123 |
+
|
| 124 |
<center> This is a demo of product recommendation using image similarity of user preferences. </center>
|
| 125 |
|
| 126 |
The the user selects their favorite product which then gets added to the user preference group. Each of the image embeddings in the user preference products get averaged into a preference embedding. Each round some products are displayed: 5 products most similar to user preference embedding and 5 random products. Embeddings are generated with [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224). The dataset used is [tonyassi/finesse1-embeddings](https://huggingface.co/datasets/tonyassi/finesse1-embeddings).
|