Distribution-Balanced Loss for Multi-Label Classification in Long-Tailed Datasets
This work solves the problem of handling imbalanced and co-occurring labels in multi-label recognition for computer vision applications, representing an incremental improvement over existing methods.
The paper tackles multi-label classification in long-tailed datasets by introducing a new loss function that addresses label co-occurrence and negative label dominance, achieving significant performance gains on Pascal VOC and COCO benchmarks.
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: https://github.com/wutong16/DistributionBalancedLoss .