发表机构
Indian Institute of Technology, Kanpur（印度理工学院坎普尔分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出 $TCP_α$ 置信度目标，通过边际控制惩罚解决现有置信度估计的歧义问题，在音乐信息检索的 rāga 识别等任务中提升了失败预测性能与域偏移鲁棒性。
AI 中文摘要
深度神经网络通常过于自信，即使对于错误预测也会赋予高置信度，导致用户缺乏可靠信号来判断何时可以信任预测结果。事后置信度估计通过在冻结分类器上训练轻量级辅助头来解决该问题，但现有目标存在固有歧义：它们为正确和错误预测分配重叠的置信度值，而决策边界附近的错误会获得与正确预测无法区分的置信度分数。本研究提出 $TCP_α$，一种新颖的置信度目标，通过为错误分类样本引入边际控制惩罚来解决这些限制。我们证明 $TCP_α$ 保证正确和错误预测的目标值完全分离，且分离边际与类别数量无关，随惩罚参数单调递增。由于准确分类器自然产生极少错误，学习这些目标会导致严重不平衡的回归问题，因此我们对不平衡情况下的训练策略进行系统研究，并通过大量 ablation 研究确定了有效的训练配置。我们在 rāga 识别任务上评估了所提方法，研究了其在域偏移下的鲁棒性，并在逐帧装饰音检测任务上进一步验证，且未修改选定的配置。在所有设置中，$TCP_α$ 在失败预测任务上始终优于现有置信度目标；仅拒绝置信度最低的 8% 的预测可将基础模型的 macro-F1 从 0.89 提升至 0.98，而仅使用来自新语料库的 5% 标注样本微调置信度头可有效恢复域偏移下的性能。
英文摘要
Deep neural networks are often overconfident, assigning high confidence even to incorrect predictions. Consequently, users lack a reliable signal for deciding when a prediction can be trusted. Post-hoc confidence estimation addresses this by training a lightweight auxiliary head over a frozen classifier. Existing targets, however, suffer from inherent ambiguity: they assign overlapping confidence values to correct and incorrect predictions, while errors near the decision boundary receive confidence scores indistinguishable from correct predictions. In this work, we propose $TCP_α$, a novel confidence target that resolves these limitations by introducing a margin-controlled penalty for misclassified samples. We prove that $TCP_α$ guarantees complete separation between the target values of correct and incorrect predictions, with a separation margin that is independent of the number of classes and increases monotonically with the penalty parameter. Since accurate classifiers naturally produce very few errors, learning these targets results in a severely imbalanced regression problem. We therefore present a systematic study of training strategies for learning under this imbalance and identify an effective training configuration through extensive ablation studies. We evaluate the proposed approach on rāga identification, investigate its robustness under domain shift, and further validate it on frame-wise ornamentation detection without modifying the selected configuration. Across all settings, $TCP_α$ consistently outperforms existing confidence targets for failure prediction. Rejecting only the least-confident 8\% of predictions improves the base model's macro-F1 from 0.89 to 0.98, while fine-tuning the confidence head with only 5\% labeled samples from a new corpus effectively restores performance under domain shift.
