# |
Team |
Task |
Date/Time |
DataID |
AMFM |
Method
|
Other Resources
|
System Description |
unuse |
unuse |
unuse |
unuse |
unuse |
unuse |
unuse |
unuse |
unuse |
unuse |
|
1 | 683 | MMEVMMen-hi | 2019/08/08 19:27:42 | 3271 | - | - | - | - | - | - | 0.770860 | - | - | - | NMT | No | This system is using VGG19 model to extract features and translated by OpenNMT tool. |
2 | NITSNLP | MMEVMMen-hi | 2019/08/10 03:27:50 | 3288 | - | - | - | - | - | - | 0.682060 | - | - | - | NMT | No | Using VGG19 with OpenNMT tool |
3 | PUP-IND | MMEVMMen-hi | 2019/08/10 16:20:01 | 3295 | - | - | - | - | - | - | 0.707520 | - | - | - | NMT | Yes | NMT with image's global features as input and local feature used for attention. Uses pre-trained embedding for English and Hindi.
|
4 | PUP-IND | MMEVMMen-hi | 2019/08/10 16:23:24 | 3296 | - | - | - | - | - | - | 0.722110 | - | - | - | NMT | Yes | This system uses various NMT models and rerank the output of models to select best candidate. Re ranking uses, Eng-Hin Dictionary, and (tri/bi)-gram model to produce score. |