Name
Abstract | ||
|---|---|---|
In this paper, we investigate the cause of the high false positive rate in Visual Relationship Detection (VRD). We observe that during training, the relationship proposal distribution is highly imbalanced: most of the negative relationship proposals are easy to identify, e.g., the inaccurate object detection, which leads to the under-fitting of low-frequency difficult proposals. This paper presents Spatially-Aware Balanced negative pRoposal sAmpling (SABRA), a robust VRD framework that alleviates the influence of false positives. To effectively optimize the model under imbalanced distribution, SABRA adopts Balanced Negative Proposal Sampling (BNPS) strategy for mini-batch sampling. BNPS divides proposals into 5 well defined sub-classes and generates a balanced training distribution according to the inverse frequency. BNPS gives an easier optimization landscape and significantly reduces the number of false positives. To further resolve the low-frequency challenging false positive proposals with high spatial ambiguity, we improve the spatial modeling ability of SABRA on two aspects: a simple and efficient multi-head heterogeneous graph attention network (MH-GAT) that models the global spatial interactions of objects, and a spatial mask decoder that learns the local spatial configuration. SABRA outperforms SOTA methods by a large margin on two human-object interaction (HOI) datasets and one general VRD dataset. |
Year | Venue | DocType |
|---|---|---|
2021 | British Machine Vision Conference | Conference |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
11 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Daisheng Jin | 1 | 0 | 0.34 |
| Xiao Ma | 2 | 1 | 2.41 |
| Chongzhi Zhang | 3 | 2 | 1.04 |
| Yizhuo Zhou | 4 | 0 | 0.34 |
| Jiashu Tao | 5 | 0 | 0.34 |
| Mingyuan Zhang | 6 | 0 | 1.01 |
| Haiyu Zhao | 7 | 65 | 6.28 |
| Shuai Yi | 8 | 167 | 14.21 |
| Zhoujun Li | 9 | 964 | 115.99 |
| Xianglong Li | 10 | 0 | 0.68 |
| Hongsheng Li | 11 | 1516 | 85.29 |