Image Captioning on Screen2Words (test)
127.9CIDErPaLI-X
Evaluation results
| Method | Links | |
|---|---|---|
| PaLI-XOCR pipeline input=false2023.05 | 127.9 | |
| Method [27]OCR pipeline input=false2023.05 | 109.4 | |
| Pix2Struct LargePretraining=Screenshot parsing, Pixel only=true2022.10 | 109.4 | |
| Pix2Struct BasePretraining=Screenshot parsing, Pixel only=true2022.10 | 107 | |
| VUTtype=State of the art w/ pipelines2022.10 | 64.3 | |
| DonutPretraining=OCR, Pixel only=true2022.10 | 56.4 |