Multi-view 3D Reconstruction on ShapeNet_R2N2 13 categories (test)
0.642Reconstruction Score (1 view)Baser2n2-AttSets (Ours)
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Baser2n2-AttSets (Ours)Backbone=Baser2n2, Aggregation=AttSets, Stage 2 training images=8, Protocol=FASet2018.08 | 0.642 | 0.662 | 0.671 | 0.676 | 0.678 | 0.686 | 0.688 | 0.693 | 0.694 | 0.694 | |
| Baser2n2-mean poolingBackbone=Baser2n2, Aggregation=mean pooling, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.632 | 0.657 | 0.665 | 0.67 | 0.672 | 0.679 | 0.681 | 0.685 | 0.686 | 0.686 | |
| Baser2n2-GRUBackbone=Baser2n2, Aggregation=GRU, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.58 | 0.616 | 0.629 | 0.637 | 0.641 | 0.649 | 0.652 | 0.652 | 0.652 | 0.652 | |
| Baser2n2-sum poolingBackbone=Baser2n2, Aggregation=sum pooling, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.58 | 0.628 | 0.644 | 0.656 | 0.66 | 0.672 | 0.677 | 0.682 | 0.684 | 0.685 | |
| Baser2n2-MHBN poolingBackbone=Baser2n2, Aggregation=MHBN pooling, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.57 | 0.596 | 0.606 | 0.612 | 0.614 | 0.621 | 0.624 | 0.628 | 0.629 | 0.629 | |
| Baser2n2-SMSO poolingBackbone=Baser2n2, Aggregation=SMSO pooling, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.57 | 0.621 | 0.641 | 0.652 | 0.656 | 0.668 | 0.673 | 0.679 | 0.68 | 0.681 | |
| Baser2n2-BP poolingBackbone=Baser2n2, Aggregation=BP pooling, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.544 | 0.599 | 0.618 | 0.628 | 0.632 | 0.644 | 0.648 | 0.654 | 0.655 | 0.656 | |
| Baser2n2-max poolingBackbone=Baser2n2, Aggregation=max pooling, Stage 2 training images=8, Protocol=Fine-tuned2018.08 | 0.524 | 0.615 | 0.641 | 0.655 | 0.661 | 0.674 | 0.678 | 0.683 | 0.684 | 0.684 |