Discriminative Triad Matching and Reconstruction for Weakly Referring Expression Grounding

Sun, Mingjie, Xiao, Jimin, Lim, Eng Gee ORCID: 0000-0003-0199-7386, Liu, Si and Goulermas, John Y
(2021) Discriminative Triad Matching and Reconstruction for Weakly Referring Expression Grounding. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 43 (11). pp. 4189-4195.

[thumbnail of manuscript.pdf] Text
manuscript.pdf - Author Accepted Manuscript

Download (2MB) | Preview


In this paper, we are tackling the weakly-supervised referring expression grounding task, for the localization of a referent object in an image according to a query sentence, where the mapping between image regions and queries are not available during the training stage. In traditional methods, an object region that best matches the referring expression is picked out, and then the query sentence is reconstructed from the selected region, where the reconstruction difference serves as the loss for back-propagation. The existing methods, however, conduct both the matching and the reconstruction approximately as they ignore the fact that the matching correctness is unknown. To overcome this limitation, a discriminative triad is designed here as the basis to the solution, through which a query can be converted into one or multiple discriminative triads in a very scalable way. Based on the discriminative triad, we further propose the triad-level matching and reconstruction modules which are lightweight yet effective for the weakly-supervised training, making it three times lighter and faster than the previous state-of-the-art methods. One important merit of our work is its superior performance despite the simple and neat design. Specifically, the proposed method achieves a new state-of-the-art accuracy when evaluated on RefCOCO (39.21 percent), RefCOCO+ (39.18 percent) and RefCOCOg (43.24 percent) datasets, that is 4.17, 4.08 and 7.8 percent higher than the previous one, respectively. The code is available at https://github.com/insomnia94/DTWREG.

Item Type: Article
Uncontrolled Keywords: Image reconstruction, Training, Proposals, Visualization, Task analysis, Linguistics, Grounding, Referring expression grounding, weakly supervised training, discriminative triad matching
Depositing User: Symplectic Admin
Date Deposited: 24 Feb 2021 15:55
Last Modified: 18 Jan 2023 22:58
DOI: 10.1109/TPAMI.2021.3058684
Related URLs:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3116000