Timezone: »

TOIST: Task Oriented Instance Segmentation Transformer with Noun-Pronoun Distillation
Pengfei Li · Beiwen Tian · Yongliang Shi · Xiaoxue Chen · Hao Zhao · Guyue Zhou · Ya-Qin Zhang

Current referring expression comprehension algorithms can effectively detect or segment objects indicated by nouns, but how to understand verb reference is still under-explored. As such, we study the challenging problem of task oriented detection, which aims to find objects that best afford an action indicated by verbs like sit comfortably on. Towards a finer localization that better serves downstream applications like robot interaction, we extend the problem into task oriented instance segmentation. A unique requirement of this task is to select preferred candidates among possible alternatives. Thus we resort to the transformer architecture which naturally models pair-wise query relationships with attention, leading to the TOIST method. In order to leverage pre-trained noun referring expression comprehension models and the fact that we can access privileged noun ground truth during training, a novel noun-pronoun distillation framework is proposed. Noun prototypes are generated in an unsupervised manner and contextual pronoun features are trained to select prototypes. As such, the network remains noun-agnostic during inference. We evaluate TOIST on the large-scale task oriented dataset COCO-Tasks and achieve +10.7% higher $\rm{mAP^{box}}$ than the best-reported results. The proposed noun-pronoun distillation can boost $\rm{mAP^{box}}$ and $\rm{mAP^{mask}}$ by +2.6% and +3.6%. Codes and models are publicly available.

Author Information

Pengfei Li (Institute for AI Industry Research, Tsinghua University)
Beiwen Tian (Tsinghua University, Tsinghua University)
Yongliang Shi (Tsinghua University, Tsinghua University)
Xiaoxue Chen (Tsinghua University, Tsinghua University)
Hao Zhao (Tsinghua University, Tsinghua University)
Guyue Zhou (Tsinghua University)
Ya-Qin Zhang (George Washington University)

More from the Same Authors