当前位置: X-MOL 学术IEEE Trans. Image Process. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Convolutional Fine-Grained Classification With Self-Supervised Target Relation Regularization
IEEE Transactions on Image Processing ( IF 10.8 ) Pub Date : 2022-08-18 , DOI: 10.1109/tip.2022.3197931
Kangjun Liu 1 , Ke Chen 2 , Kui Jia 2
Affiliation  

Fine-grained visual classification can be addressed by deep representation learning under supervision of manually pre-defined targets ( e.g., one-hot or the Hadamard codes). Such target coding schemes are less flexible to model inter-class correlation and are sensitive to sparse and imbalanced data distribution as well. In light of this, this paper introduces a novel target coding scheme – dynamic target relation graphs (DTRG), which, as an auxiliary feature regularization, is a self-generated structural output to be mapped from input images. Specifically, online computation of class-level feature centers is designed to generate cross-category distance in the representation space, which can thus be depicted by a dynamic graph in a non-parametric manner. Explicitly minimizing intra-class feature variations anchored on those class-level centers can encourage learning of discriminative features. Moreover, owing to exploiting inter-class dependency, the proposed target graphs can alleviate data sparsity and imbalanceness in representation learning. Inspired by recent success of the mixup style data augmentation, this paper introduces randomness into soft construction of dynamic target relation graphs to further explore relation diversity of target classes. Experimental results can demonstrate the effectiveness of our method on a number of diverse benchmarks of multiple visual classification, especially achieving the state-of-the-art performance on three popular fine-grained object benchmarks and superior robustness against sparse and imbalanced data. Source codes are made publicly available at https://github.com/AkonLau/DTRG .

中文翻译:

带有自监督目标关系正则化的卷积细粒度分类

细粒度的视觉分类可以通过在手动预定义目标的监督下进行深度表示学习来解决( 例如,one-hot 或 Hadamard 代码)。这种目标编码方案对类间相关性建模的灵活性较低,并且对稀疏和不平衡的数据分布也很敏感。鉴于此,本文介绍了一种新颖的目标编码方案——动态目标关系图(DTRG),它作为辅助特征正则化,是从输入图像映射的自生成结构输出。具体来说,类级特征中心的在线计算旨在生成表示空间中的跨类别距离,因此可以通过动态图以非参数方式描述。显式最小化锚定在这些类级中心上的类内特征变化可以鼓励对判别特征的学习。此外,由于利用了类间依赖,所提出的目标图可以缓解表示学习中的数据稀疏性和不平衡性。受最近混合风格数据增强成功的启发,本文将随机性引入动态目标关系图的软构建中,以进一步探索目标类的关系多样性。实验结果可以证明我们的方法在多个视觉分类的多个不同基准上的有效性,特别是在三个流行的细粒度对象基准上实现了最先进的性能,并且对稀疏和不平衡数据具有出色的鲁棒性。源代码在以下位置公开提供 本文将随机性引入到动态目标关系图的软构建中,以进一步探索目标类的关系多样性。实验结果可以证明我们的方法在多个视觉分类的多个不同基准上的有效性,特别是在三个流行的细粒度对象基准上实现了最先进的性能,并且对稀疏和不平衡数据具有出色的鲁棒性。源代码在以下位置公开提供 本文将随机性引入到动态目标关系图的软构建中,以进一步探索目标类的关系多样性。实验结果可以证明我们的方法在多个视觉分类的多个不同基准上的有效性,特别是在三个流行的细粒度对象基准上实现了最先进的性能,并且对稀疏和不平衡数据具有出色的鲁棒性。源代码在以下位置公开提供 特别是在三个流行的细粒度对象基准上实现了最先进的性能,以及对稀疏和不平衡数据的卓越鲁棒性。源代码在以下位置公开提供 特别是在三个流行的细粒度对象基准上实现了最先进的性能,以及对稀疏和不平衡数据的卓越鲁棒性。源代码在以下位置公开提供https://github.com/AkonLau/DTRG .
更新日期:2022-08-18
down
wechat
bug