J.-Y. Yang

Movie Soundtrack Emotion Matching Algorithm Based on CLIP Model

To address the semantic distortion and temporal misalignment introduced by existing movie soundtrack matching methods that rely on textual intermediaries, this paper proposes an end-to-end audio-visual cross-modal temporal alignment algorithm based on the CLIP model. Considering that robust multimodal signal fusion and...