Action Quality Assessment (AQA) evaluates diverse skills but models struggle with non-stationary data. We propose Continual AQA (CAQA) to refine models using sparse new data. Feature replay preserves memory without storing raw inputs. However, the misalignment between static old features and the dynamically changing feature manifold causes severe catastrophic forgetting. To address this novel problem, we propose Manifold-Aligned Graph Regularization (MAGR), which first aligns deviated old features to the current feature manifold, ensuring representation consistency. It then constructs a graph jointly arranging old and new features aligned with quality scores. Experiments show MAGR outperforms recent strong baselines with up to 6.56%, 5.66%, 15.64%, and 9.05% correlation gains on the MTL-AQA, FineDiving, UNLV-Dive, and JDM-MSA split datasets, respectively. This validates MAGR for continual assessment challenges arising from non-stationary skill variations.

翻译：动作质量评估（AQA）涉及对多样化技能进行评价，但现有模型在处理非平稳数据时面临挑战。我们提出持续动作质量评估（CAQA）框架，旨在通过稀疏的新数据精化模型。特征回放机制可在不存储原始输入的前提下保留记忆。然而，静态旧特征与动态变化的特征流形之间的错位会引发严重的灾难性遗忘。为解决这一新问题，我们提出流形对齐图正则化（MAGR）方法：首先将偏移的旧特征对齐至当前特征流形，确保表征一致性；继而构建联合排列新旧特征并关联质量分数的图结构。实验结果表明，在MTL-AQA、FineDiving、UNLV-Dive和JDM-MSA分割数据集上，MAGR相较于近期强基线方法分别取得6.56%、5.66%、15.64%和9.05%的相关性增益。这验证了MAGR在应对非平稳技能变化引发的持续评估挑战中的有效性。