Commit
·
5043278
1
Parent(s):
3c44d8a
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
|
| 2 |
# Visual semantic with BERT-CNN
|
| 3 |
|
| 4 |
-
This model can be used to assign an object-to-caption semantic
|
| 5 |
(1) caption diverse re-ranking, and (2) generate soft labels for caption filtering when scraping text-to-captions from the internet.
|
| 6 |
|
| 7 |
To take advantage of the overlapping between the visual context and the caption, and to extract global information from each visual (i.e., object, scene, etc) we use BERT as an embedding layer followed by a shallow CNN (tri-gram kernel) (Kim, 2014).
|
|
@@ -333,7 +333,6 @@ class BertLayer(tf.keras.layers.Layer):
|
|
| 333 |
df = pd.read_csv(args.train, sep='\t')
|
| 334 |
|
| 335 |
|
| 336 |
-
#labels = df.is_duplicate.values
|
| 337 |
labels = df.is_related.values
|
| 338 |
|
| 339 |
texts = []
|
|
|
|
| 1 |
|
| 2 |
# Visual semantic with BERT-CNN
|
| 3 |
|
| 4 |
+
This model can be used to assign an object-to-caption semantic ness score, which is valuable for
|
| 5 |
(1) caption diverse re-ranking, and (2) generate soft labels for caption filtering when scraping text-to-captions from the internet.
|
| 6 |
|
| 7 |
To take advantage of the overlapping between the visual context and the caption, and to extract global information from each visual (i.e., object, scene, etc) we use BERT as an embedding layer followed by a shallow CNN (tri-gram kernel) (Kim, 2014).
|
|
|
|
| 333 |
df = pd.read_csv(args.train, sep='\t')
|
| 334 |
|
| 335 |
|
|
|
|
| 336 |
labels = df.is_related.values
|
| 337 |
|
| 338 |
texts = []
|