Displayed are triplets annotated with human referential utterances. A neural listener, trained with chairs, is used to color-code attention-wise the utterances and score each object. Underlined tokens where out-of-the-vocabulary and were ignored by the listener.
| Distractor-A | Distractor-B | Target |
|
|
|
|
Human Speaker:
the bed has a headboard and footboard and Correctly guessed: True Confidences: 0.00, 0.01, 0.98 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
|
Human Speaker: wavy Correctly guessed: True Confidences: 0.00, 0.01, 0.99 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
Human Speaker: the bed is square in shape Correctly guessed: True Confidences: 0.01, 0.00, 0.99 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
|
Human Speaker:
the bed is square shaped and the headboard Correctly guessed: True Confidences: 0.01, 0.00, 0.99 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
Human Speaker: wrinkly sheets Correctly guessed: True Confidences: 0.06, 0.11, 0.83 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
|
Human Speaker:
it 's the bed with the wrinkly looking comforter Correctly guessed: True Confidences: 0.01, 0.01, 0.98 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
Human Speaker:
no footboard head board is tall with slats Correctly guessed: True Confidences: 0.25, 0.03, 0.72 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
|
Human Speaker: the headboard has thin horizontal bars Correctly guessed: True Confidences: 0.02, 0.00, 0.98 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
|
Human Speaker: many vertical slots Correctly guessed: True Confidences: 0.01, 0.01, 0.98 |
||
| Distractor-A | Distractor-B | Target |
|
|
|
|
|
Human Speaker:
this bed has one pillow and the bed frame Correctly guessed: True Confidences: 0.01, 0.02, 0.98 |
||