Abstract:Modern visual recognition models often display overconfidence due to their reliance on complex deep neural networks and one-hot target supervision, resulting in unreliable confidence scores that necessitate calibration. While current confidence calibration techniques primarily address single-label scenarios, there is a lack of focus on more practical and generalizable multi-label contexts. This paper introduces the Multi-Label Confidence Calibration (MLCC) task, aiming to provide well-calibrated confidence scores in multi-label scenarios. Unlike single-label images, multi-label images contain multiple objects, leading to semantic confusion and further unreliability in confidence scores. Existing single-label calibration methods, based on label smoothing, fail to account for category correlations, which are crucial for addressing semantic confusion, thereby yielding sub-optimal performance. To overcome these limitations, we propose the Dynamic Correlation Learning and Regularization (DCLR) algorithm, which leverages multi-grained semantic correlations to better model semantic confusion for adaptive regularization. DCLR learns dynamic instance-level and prototype-level similarities specific to each category, using these to measure semantic correlations across different categories. With this understanding, we construct adaptive label vectors that assign higher values to categories with strong correlations, thereby facilitating more effective regularization. We establish an evaluation benchmark, re-implementing several advanced confidence calibration algorithms and applying them to leading multi-label recognition (MLR) models for fair comparison. Through extensive experiments, we demonstrate the superior performance of DCLR over existing methods in providing reliable confidence scores in multi-label scenarios.

Confidence Self-Calibration for Multi-Label Class-Incremental Learning

Rebalancing Multi-Label Class-Incremental Learning

FGCM: Noisy Label Learning via Fine-Grained Confidence Modeling

Dynamic Correlation Learning and Regularization for Multi-Label Confidence Calibration

Overcoming Catastrophic Forgetting for Multi-Label Class-Incremental Learning

Class-Incremental Lifelong Learning in Multi-Label Classification

Knowledge Restore and Transfer for Multi-label Class-Incremental Learning

Confidence-Aware Calibration and Scoring Functions for Curriculum Learning

Partial Multi-Label Learning with Mutual Teaching

Calibrating Multimodal Learning

Calibrated Multi-label Classification with Label Correlations

Multi-Label Continual Learning Using Augmented Graph Convolutional Network

M2SD:Multiple Mixing Self-Distillation for Few-Shot Class-Incremental Learning

Combining Past, Present and Future: A Self-Supervised Approach for Class Incremental Learning.

NeCa: Network Calibration for Class Incremental Learning.

Uncertainty-Guided Semi-Supervised Few-Shot Class-Incremental Learning With Knowledge Distillation

Double Confidence Calibration Focused Distillation for Task-Incremental Learning

Calibrate to Discriminate: Improve In-Context Learning with Label-Free Comparative Inference

Confident Learning: Estimating Uncertainty in Dataset Labels

More Classifiers, Less Forgetting: A Generic Multi-classifier Paradigm for Incremental Learning

Max-margin Class Incremental Learning with Mixup Augmentation