Cross-modal relevance matching on Commercial Advertising Dataset (test)
91.73AUCAlignCMSS
Evaluation Results
| Method | Links | |
|---|---|---|
| AlignCMSSAds Input=Multi-Modal, Training Data Percentage=100%2023.09 | 91.73 | |
| AlignCMSSAds Input=Multi-Modal, Training Data Percentage=70%2023.09 | 91.02 | |
| OSCARAds Input=Multi-Modal, VALSE Integration=true2023.09 | 90.62 | |
| AlignCMSSAds Input=Multi-Modal, Training Data Percentage=50%2023.09 | 90.46 | |
| VL-BERTAds Input=Multi-Modal, VALSE Integration=true2023.09 | 90.13 | |
| AdsCVLRAds Input=Multi-Modal2023.09 | 89.16 | |
| VinVLAds Input=Multi-Modal2023.09 | 88.56 | |
| Unicoder-VL@(Qt-OtOi)Ads Input=Multi-Modal2023.09 | 87.9 | |
| OSCARAds Input=Multi-Modal, VALSE Integration=false2023.09 | 87.45 | |
| VL-BERTAds Input=Multi-Modal, VALSE Integration=false2023.09 | 86.27 | |
| BLIPAds Input=Multi-Modal2023.09 | 83.51 | |
| Unicoder-VLAds Input=Image2023.09 | 83.16 | |
| ALBEFAds Input=Multi-Modal2023.09 | 82.74 | |
| ViLTAds Input=Image2023.09 | 82.23 | |
| CLIP-MH@(Qt-OtOi)Ads Input=Multi-Modal2023.09 | 82.18 | |
| BERT-baseAds Input=Text2023.09 | 82.15 | |
| CLIP@(Qt-OtOi)Ads Input=Multi-Modal2023.09 | 81.98 | |
| CLIPAds Input=Image2023.09 | 81.16 |