Referring Expression Segmentation on Reason-Seg (test)
90.4gIoUEmpirical Upper Bound
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Empirical Upper BoundTraining regime=Oracle reference2026.05 | 90.4 | 90.6 | |
| SAM-Veteran-7BTraining regime=Reinforcement-Learning Tuning2026.05 | 62.6 | 56.1 | |
| PinPoint_multTraining regime=Training-Free2026.05 | 61.7 | 55.4 | |
| LISA-13B-LLaVA1.5Finetune=true2024.06 | 61.3 | 62.2 | |
| PinPoint_msmTraining regime=Training-Free2026.05 | 60.9 | 54.9 | |
| SAM-R1-7BTraining regime=Reinforcement-Learning Tuning2026.05 | 60.2 | 54.3 | |
| Seg-Zero-7BTraining regime=Reinforcement-Learning Tuning2026.05 | 57.5 | 52 | |
| LISA-7B-LLaVA1.5Finetune=true2024.06 | 55.6 | 56.9 | |
| EVF-SAMFinetune=true2024.06 | 55.1 | 50.5 | |
| LISA-7B-LLaVA1.5Finetune=false2024.06 | 48.7 | 48.8 | |
| LISA-7BTraining regime=Supervised Fine-tuning2026.05 | 48.7 | 48.8 | |
| EVF-SAMFinetune=false2024.06 | 38.9 | 32.6 | |
| LISA-7B-LLaVA1.1Finetune=false2024.06 | 36.8 | 34.1 | |
| LISA-7B-LLaVA1.1Finetune=true2024.06 | 36.8 | 34.1 | |
| OVSegFinetune=false2024.06 | 26.1 | 20.8 | |
| SEEMFinetune=false2024.06 | 24.3 | 18.7 | |
| X-DecoderFinetune=false2024.06 | 21.7 | 16.3 |