EXPLORE CONNECTIONS
Optimizing VLP-aligned Multimodal Intent Representation with Correct Visual Instantiation for Zero-Shot Composed Image Retrieval
Follow the relationships that help you find your next source.
Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.