LIMSSR: LLM-Driven Sequence-to-Score Reasoning under Training-Time Incomplete Multimodal Observations
For practitioners of multimodal learning facing missing data, this work provides a practical solution that eliminates the unrealistic assumption of complete training data, enabling more robust real-world applications.
LIMSSR addresses incomplete multimodal learning under training-time missing modalities, using LLM-driven reasoning to infer semantics without complete data. It achieves state-of-the-art results on three Action Quality Assessment datasets, outperforming baselines without requiring full-modal training data.
Real-world multimodal learning is often hindered by missing modalities. While Incomplete Multimodal Learning (IML) has gained traction, existing methods typically rely on the unrealistic assumption of full-modal availability during training to provide reconstruction supervision or cross-modal priors. This paper tackles the more challenging setting of IML under training-time incomplete observations, which precludes reliance on a ``God's eye view'' of complete data. We propose LIMSSR (LLM-Driven Incomplete Multimodal Sequence-to-Score Reasoning), a framework that reformulates this challenge as a conditional sequence reasoning task. LIMSSR leverages the semantic reasoning capabilities of Large Language Models via Prompt-Guided Context-Aware Modality Imputation and Multidimensional Representation Fusion to infer latent semantics from available contexts without direct reconstruction. To mitigate hallucinations, we introduce a Mask-Aware Dual-Path Aggregation to dynamically calibrate inference uncertainty. Extensive experiments on three Action Quality Assessment datasets demonstrate that LIMSSR significantly outperforms state-of-the-art baselines without relying on complete training data, establishing a new paradigm for data-efficient multimodal learning. Code is available at https://github.com/XuHuangbiao/LIMSSR.