Text this: Cross-modal deepfake detection: integrating textual and frequency domains