ExpansionNet v2 (No VL pretraining)

Reported on 6 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing6 results

Image CaptioningonCOCO Captions
ROUGE-L· 2022-08-13
61.1
SOTA
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning arXiv:2208.06551
Image CaptioningonCOCO Captions
BLEU-1· 2022-08-13
83.5
best: 84.2 (GRIT (No VL pretraining - base))
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning arXiv:2208.06551
Image CaptioningonCOCO Captions
BLEU-4· 2022-08-13
42.7
best: 46.5 (mPLUG)
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning arXiv:2208.06551
Image CaptioningonCOCO Captions
CIDER· 2022-08-13
143.7
best: 155.1 (mPLUG)
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning arXiv:2208.06551
Image CaptioningonCOCO Captions
METEOR· 2022-08-13
30.6
best: 33.9 (CoCa)
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning arXiv:2208.06551
Image CaptioningonCOCO Captions
SPICE· 2022-08-13
24.7
best: 27 (VAST)
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning arXiv:2208.06551