面向无偏场景图生成的头尾协同学习网络
Head-Tail Cooperative Learning Network for Unbiased Scene Graph Generation
AI总结：
针对场景图生成中谓词长尾分布导致的头部偏置问题，提出头尾协同学习网络，通过双分支协同与自监督学习平衡头尾谓词识别，在多个数据集上以极小Recall损失实现更高平均Recall与最优整体性能。
摘要：
场景图生成（Scene Graph Generation, SGG）作为图像理解中的关键任务，面临着由谓词长尾分布导致的头部偏置预测挑战。然而，当前的无偏SGG方法往往容易优先提升尾部谓词的预测效果，却忽略了头部谓词预测性能的大幅下降，最终导致偏差从头部转向尾部。为解决这一问题，我们提出了一种模型无关的头尾协同学习（Head-Tail Collaborative Learning, HTCL）网络，该网络包含偏好头部和偏好尾部的特征表示分支，二者协同工作以实现对头部和尾部谓词的精准识别。我们还提出了一种自监督学习方法，通过约束偏好尾部的谓词特征来增强偏好尾部特征表示分支的预测能力。具体而言，该自监督学习通过对比学习和头部中心损失，使头部谓词特征向其类别中心收敛，同时尽可能分散尾部谓词特征。我们在VG150、Open Images V6和GQA200数据集上将HTCL应用于多种SGG模型，验证了其有效性。结果表明，我们的方法在Recall指标损失极小的情况下实现了更高的平均Recall，并且取得了全新的最优整体性能。我们的代码可在https://github.com/wanglei0618/HTCL 获取。
英文摘要：
Scene Graph Generation (SGG) as a critical task in image understanding, facing the challenge of head-biased prediction caused by the long-tail distribution of predicates. However, current unbiased SGG methods can easily prioritize improving the prediction of tail predicates while ignoring the substantial sacrifice in the prediction of head predicates, leading to a shift from head bias to tail bias. To address this issue, we propose a model-agnostic Head-Tail Collaborative Learning (HTCL) network that includes head-prefer and tail-prefer feature representation branches that collaborate to achieve accurate recognition of both head and tail predicates. We also propose a self-supervised learning approach to enhance the prediction ability of the tail-prefer feature representation branch by constraining tail-prefer predicate features. Specifically, self-supervised learning converges head predicate features to their class centers while dispersing tail predicate features as much as possible through contrast learning and head center loss. We demonstrate the effectiveness of our HTCL by applying it to various SGG models on VG150, Open Images V6 and GQA200 datasets. The results show that our method achieves higher mean Recall with a minimal sacrifice in Recall and achieves a new state-of-the-art overall performance. Our code is available at https://github.com/wanglei0618/HTCL.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
