Abstract:Conventional detectors suffer from performance degradation when dealing with long-tailed data due to a classification bias towards the majority head categories. In this paper, we contend that the learning bias originates from two factors: 1) the unequal competition arising from the imbalanced distribution of foreground categories, and 2) the lack of sample diversity in tail categories. To tackle these issues, we introduce a unified framework called BAlanced CLassification (BACL), which enables adaptive rectification of inequalities caused by disparities in category distribution and dynamic intensification of sample diversities in a synchronized manner. Specifically, a novel foreground classification balance loss (FCBL) is developed to ameliorate the domination of head categories and shift attention to difficult-to-differentiate categories by introducing pairwise class-aware margins and auto-adjusted weight terms, respectively. This loss prevents the over-suppression of tail categories in the context of unequal competition. Moreover, we propose a dynamic feature hallucination module (FHM), which enhances the representation of tail categories in the feature space by synthesizing hallucinated samples to introduce additional data variances. In this divide-and-conquer approach, BACL sets a new state-of-the-art on the challenging LVIS benchmark with a decoupled training pipeline, surpassing vanilla Faster R-CNN with ResNet-50-FPN by 5.8% AP and 16.1% AP for overall and tail categories. Extensive experiments demonstrate that BACL consistently achieves performance improvements across various datasets with different backbones and architectures. Code and models are available at <a class="link-external link-https" href="https://github.com/Tianhao-Qi/BACL" rel="external noopener nofollow">this https URL</a>.

GREB: Gradient Re-Balanced Loss for Long-Tailed Multi-Lable Classification

Balanced complement loss for long-tailed image classification

Gradient-Aware Logit Adjustment Loss for Long-tailed Classifier

The Equalization Losses: Gradient-Driven Training for Long-tailed Object Recognition

Long-Tailed Learning as Multi-Objective Optimization

Balanced Gradient Penalty Improves Deep Long-Tailed Learning

Balanced Contrastive Learning for Long-Tailed Visual Recognition

A Dual-Branch Learning Model with Gradient-Balanced Loss for Long-Tailed Multi-Label Text Classification

Feature Re-Balancing for Long-Tailed Visual Recognition.

Long-tailed Visual Recognition via Gaussian Clouded Logit Adjustment

Adaptively Weighted Copy-Decoupling Resampling Strategy for Long-Tailed Multi-label Classification

Decoupling Representation and Classifier for Long-Tailed Recognition

Enhanced multi-branch learning for long-tailed image recognition

Long-Tailed Recognition by Hierarchical Rebalancing Dual-Classifier

Feature-Balanced Loss for Long-Tailed Visual Recognition

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

Balanced Classification: A Unified Framework for Long-Tailed Object Detection

Long-Tailed Classification Based on Coarse-Grained Leading Forest and Multi-Center Loss

Rectify the Regression Bias in Long-Tailed Object Detection

Difficulty-aware Balancing Margin Loss for Long-tailed Recognition

Compound Batch Normalization for Long-tailed Image Classification