All Stories

  1. CRISPR: Context-Refined Information Spatial Pooling with Region-awareness for Efficient Visual Token Compression in VLMs
  2. Few-Shot Multimodal Explanation for Visual Question Answering
  3. Semantic Editing Increment Benefits Zero-Shot Composed Image Retrieval
  4. LDRE: LLM-based Divergent Reasoning and Ensemble for Zero-Shot Composed Image Retrieval
  5. Open-World Social Event Classification
  6. MMT: Image-guided Story Ending Generation with Multimodal Memory Transformer