Referring Video Object Segmentation on DAVIS 2017 (val)
45.3J&F ScoreBellver et al. (2020) + SynthRef
Evaluation results
| Method | Links | |
|---|---|---|
| Bellver et al. (2020) + SynthRefSynthRef=true, Pre-training=RefCOCO + SynthRef-YouTube-VIS, Fine-tuning=DAVIS-2017 training data2021.06 | 45.3 | |
| Bellver et al. (2020)SynthRef=false, Pre-training=RefCOCO, Fine-tuning=DAVIS-2017 training data2021.06 | 45.1 | |
| Seo et al. (2020)SynthRef=false, Pre-training=RefCOCO, Fine-tuning=DAVIS-2017 training data2021.06 | 44.1 | |
| Khoreva et al. (2018)SynthRef=false, Pre-training=RefCOCO, Fine-tuning=DAVIS-2017 training data2021.06 | 39.3 |