Counting on Robotic Counting (val)
100AccuracyEmbodied Counting model
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Embodied Counting modelModel ID=E3, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=100%2026.04 | 100 | 1.05 | |
| Embodied Counting modelModel ID=E6, Pretrain=No, Joint=Yes, Curriculum=Random, Data Fraction=100%2026.04 | 100 | 1.71 | |
| Embodied Counting modelModel ID=E9, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=100%2026.04 | 100 | 0.74 | |
| Embodied Counting modelModel ID=E12, Pretrain=Yes, Joint=Yes, Curriculum=Random, Data Fraction=100%2026.04 | 100 | 1.39 | |
| Embodied Counting modelModel ID=E3, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=50%2026.04 | 99.9 | 1.68 | |
| Embodied Counting modelModel ID=E6, Pretrain=No, Joint=Yes, Curriculum=Random, Data Fraction=50%2026.04 | 99.9 | 2.21 | |
| Embodied Counting modelModel ID=E9, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=50%2026.04 | 99.9 | 0.69 | |
| Embodied Counting modelModel ID=E12, Pretrain=Yes, Joint=Yes, Curriculum=Random, Data Fraction=50%2026.04 | 99.8 | 1.35 | |
| Embodied Counting modelModel ID=E10, Pretrain=Yes, Joint=Yes, Curriculum=Easy→Hard, Data Fraction=50%2026.04 | 99.2 | 1.87 | |
| Embodied Counting modelModel ID=E7, Pretrain=Yes, Joint=No, Curriculum=Easy→Hard, Data Fraction=50%2026.04 | 99.1 | 1.39 | |
| Embodied Counting modelModel ID=E7, Pretrain=Yes, Joint=No, Curriculum=Easy→Hard, Data Fraction=100%2026.04 | 99.1 | 1.33 | |
| Embodied Counting modelModel ID=E10, Pretrain=Yes, Joint=Yes, Curriculum=Easy→Hard, Data Fraction=100%2026.04 | 99.1 | 1.69 | |
| Embodied Counting modelModel ID=E8, Pretrain=Yes, Joint=No, Curriculum=Hard→Easy, Data Fraction=50%2026.04 | 98.6 | 1.08 | |
| Embodied Counting modelModel ID=E11, Pretrain=Yes, Joint=Yes, Curriculum=Hard→Easy, Data Fraction=50%2026.04 | 98.5 | 1.59 | |
| Vision-only baselineModel ID=V1, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=100%2026.04 | 97.6 | — | |
| Embodied Counting modelModel ID=E1, Pretrain=No, Joint=No, Curriculum=Easy→Hard, Data Fraction=50%2026.04 | 97.2 | 1.54 | |
| Embodied Counting modelModel ID=E4, Pretrain=No, Joint=Yes, Curriculum=Easy→Hard, Data Fraction=50%2026.04 | 97.2 | 1.97 | |
| Embodied Counting modelModel ID=E4, Pretrain=No, Joint=Yes, Curriculum=Easy→Hard, Data Fraction=100%2026.04 | 97.2 | 2.19 | |
| Embodied Counting modelModel ID=E9, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=10%2026.04 | 96.8 | 1.46 | |
| Embodied Counting modelModel ID=E12, Pretrain=Yes, Joint=Yes, Curriculum=Random, Data Fraction=10%2026.04 | 96.6 | 2.02 | |
| Embodied Counting modelModel ID=E2, Pretrain=No, Joint=No, Curriculum=Hard→Easy, Data Fraction=50%2026.04 | 96.3 | 1.67 | |
| Vision-only baselineModel ID=V2, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=100%2026.04 | 96.3 | — | |
| Embodied Counting modelModel ID=E8, Pretrain=Yes, Joint=No, Curriculum=Hard→Easy, Data Fraction=100%2026.04 | 96.3 | 1.03 | |
| Embodied Counting modelModel ID=E1, Pretrain=No, Joint=No, Curriculum=Easy→Hard, Data Fraction=100%2026.04 | 95.8 | 1.77 | |
| Vision-only baselineModel ID=V2, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=50%2026.04 | 94.8 | — | |
| Vision-only baselineModel ID=V1, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=50%2026.04 | 94.7 | — | |
| Embodied Counting modelModel ID=E7, Pretrain=Yes, Joint=No, Curriculum=Easy→Hard, Data Fraction=10%2026.04 | 94.7 | 1.42 | |
| Embodied Counting modelModel ID=E10, Pretrain=Yes, Joint=Yes, Curriculum=Easy→Hard, Data Fraction=10%2026.04 | 94.3 | 1.95 | |
| Embodied Counting modelModel ID=E2, Pretrain=No, Joint=No, Curriculum=Hard→Easy, Data Fraction=100%2026.04 | 94.2 | 1.87 | |
| Embodied Counting modelModel ID=E8, Pretrain=Yes, Joint=No, Curriculum=Hard→Easy, Data Fraction=10%2026.04 | 94.1 | 1.46 | |
| Embodied Counting modelModel ID=E11, Pretrain=Yes, Joint=Yes, Curriculum=Hard→Easy, Data Fraction=10%2026.04 | 94 | 1.97 | |
| Vision-only baselineModel ID=V4, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=100%2026.04 | 91.9 | — | |
| Vision-only baselineModel ID=V3, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=100%2026.04 | 91.6 | — | |
| Embodied Counting modelModel ID=E6, Pretrain=No, Joint=Yes, Curriculum=Random, Data Fraction=10%2026.04 | 90.2 | 2.08 | |
| Embodied Counting modelModel ID=E3, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=10%2026.04 | 90.1 | 1.7 | |
| Embodied Counting modelModel ID=E11, Pretrain=Yes, Joint=Yes, Curriculum=Hard→Easy, Data Fraction=100%2026.04 | 85.9 | 1.9 | |
| Embodied Counting modelModel ID=E1, Pretrain=No, Joint=No, Curriculum=Easy→Hard, Data Fraction=10%2026.04 | 85.7 | 1.73 | |
| Embodied Counting modelModel ID=E4, Pretrain=No, Joint=Yes, Curriculum=Easy→Hard, Data Fraction=10%2026.04 | 85.5 | 2.05 | |
| Embodied Counting modelModel ID=E5, Pretrain=No, Joint=Yes, Curriculum=Hard→Easy, Data Fraction=50%2026.04 | 85.2 | 2.11 | |
| Embodied Counting modelModel ID=E2, Pretrain=No, Joint=No, Curriculum=Hard→Easy, Data Fraction=10%2026.04 | 84 | 1.74 | |
| Embodied Counting modelModel ID=E5, Pretrain=No, Joint=Yes, Curriculum=Hard→Easy, Data Fraction=10%2026.04 | 84 | 2.1 | |
| Vision-only baselineModel ID=V3, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=50%2026.04 | 83.3 | — | |
| Embodied Counting modelModel ID=E5, Pretrain=No, Joint=Yes, Curriculum=Hard→Easy, Data Fraction=100%2026.04 | 82.1 | 2.31 | |
| Vision-only baselineModel ID=V4, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=50%2026.04 | 81.8 | — | |
| Vision-only baselineModel ID=V2, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=10%2026.04 | 77.9 | — | |
| Vision-only baselineModel ID=V1, Pretrain=No, Joint=No, Curriculum=Random, Data Fraction=10%2026.04 | 75.4 | — | |
| Vision-only baselineModel ID=V4, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=10%2026.04 | 61 | — | |
| Vision-only baselineModel ID=V3, Pretrain=Yes, Joint=No, Curriculum=Random, Data Fraction=10%2026.04 | 60.6 | — |