AhmedSSabir
commited on
Commit
•
e40d0bf
1
Parent(s):
5e7390c
Update README.md
Browse files
README.md
CHANGED
@@ -1,8 +1,9 @@
|
|
1 |
|
2 |
# Visual semantic with BERT-CNN
|
3 |
|
4 |
-
This model can be used to assign an object-to-caption semantic relatedness score, which is valuable for
|
5 |
-
|
|
|
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).
|
8 |
|
|
|
1 |
|
2 |
# Visual semantic with BERT-CNN
|
3 |
|
4 |
+
This model can be used to assign an object-to-caption semantic relatedness score, which is valuable for (1) caption diverse re-ranking (this work),
|
5 |
+
and (2) (as an application) generating soft labels for filtering image-to-caption when scraping text-to-captions from the internet (e,g., Instagram).
|
6 |
+
|
7 |
|
8 |
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).
|
9 |
|