Distribution-Balanced Loss for Multi-Label Classification in Long-Tailed Datasets

By Tong Wu, Qingqiu Huang, Ziwei Liu, Yu Wang, Dahua Lin. European Conference on Computer Vision, 2020

299 citations recorded by the source corpus.

We present a new loss function called Distribution-Balanced Loss for the multi-label recognition problems that exhibit long-tailed class distributions. Compared to conventional single-label classification problem, multi-label recognition problems are often more challenging due to two significant issues, namely the co-occurrence of labels and the dominance of negative labels (when treated as multiple binary classification problems). The Distribution-Balanced Loss tackles these issues through two key modifications to the standard binary cross-entropy loss: 1) a new way to re-balance the weights that takes into account the impact caused by label co-occurrence, and 2) a negative tolerant regularization to mitigate the over-suppression of negative labels. Experiments on both Pascal VOC and COCO show that the models trained with this new loss function achieve significant performance gains over existing methods. Code and models are available at: this https URL .

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/7eda53373b495c9b5d5a93c6371bcc6e3616171d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jiashi Feng
2. Ziwei Liu
3. Chen Change Loy
4. Dahua Lin
5. Kai Chen
6. Shuicheng Yan
7. Jiangmiao Pang
8. Bingyi Kang

Citations in the corpus

• Long-tailed Recognition by Routing Diverse Distribution-Aware Experts (2020, unknown): “…balancing, re-weighting the loss functions can also introduce learning balance by either putting larger weights on more challenging and sparse classes (Lin et al., 2017; Cui et al., 2019; Cao et al., 2019; Wu et al., 2020) or randomly ignoring gradients from head classes (Tan et al., 2020).”
• Seesaw Loss for Long-Tailed Instance Segmentation (2020, background): “Long-tailed recognition tasks [32, 14, 15, 55] receive growing attention recently as the problems are closer to real-world applications.”
• Long-Tailed Classification by Keeping the Good and Removing the Bad Momentum Causal Effect (2020, unknown): “For example, in instance segmentation [7], if we target at increasing the images of tail class instances like “remote controller”, we have to bring in more head instances like “sofa” and “TV” simultaneously in every newly added image [8].”
• Asymmetric Loss For Multi-Label Classification (2020, background): “Also notice that logits shifting, as suggested in [19] and [33], is different from probability shifting due to the nonlinear sigmoid operation.”
• Deep Long-Tailed Learning: A Survey (2021, methodology): “In multi-label image classification, the benchmarks are VOC-LT [37] and COCOLT [37], which are sampled from PASCAL VOC 2012 [42] and COCO [43], respectively.”