Research map / Vision-language and multimodal / Cross-modal learning
Cross-Modality Personalization for Retrieval
Nils Murrugarra-Llerena, Adriana Kovashka
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Combine content, style, and base embeddings | Cross-Modality Personalization for Retrieval · Ads gaze-to-text | top-3 accuracy: 0.4366 -> 0.4463 | Improved |
| Combine content, style, and base embeddings | Cross-Modality Personalization for Retrieval · Ads text-to-gaze | top-3 accuracy: 0.4384 -> 0.4426 | Improved |
| Combine content, style, and base embeddings | Cross-Modality Personalization for Retrieval · Ads text-to-personality | top-3 accuracy: 0.2894 -> 0.2861 | Hurt |
| Combine content, style, and base embeddings | Cross-Modality Personalization for Retrieval · COCO text-to-gaze | top-3 accuracy: 0.6139 -> 0.6250 | Improved |
| Combine content, style, and base embeddings | Cross-Modality Personalization for Retrieval · COCO gaze-to-text | top-3 accuracy: 0.5972 -> 0.6065 | Improved |
| Combine content, style, and base embeddings | Cross-Modality Personalization for Retrieval · COCO personality-to-gaze | top-3 accuracy: 0.1759 -> 0.1639 | Hurt |
| Add style constraints to content and base modeling (ablation) | Cross-Modality Personalization for Retrieval · Ads gaze-to-text | top-3 accuracy: 0.3745 -> 0.4463 | Improved |
| Add content constraints to style and base modeling (ablation) | Cross-Modality Personalization for Retrieval · Ads text-to-gaze | top-3 accuracy: 0.4023 -> 0.4426 | Improved |
| Use style constraints only instead of the combined embedding (ablation) | Cross-Modality Personalization for Retrieval · Ads text-to-gaze | top-3 accuracy: 0.2704 -> 0.4426 | Hurt |
| Train style constraints jointly across all modalities rather than per task (ablation) | Style embedding · Ads personality-to-gaze | top-3 accuracy: 0.1643 -> 0.1815 | Improved |
| Train style constraints jointly across all modalities rather than per task (ablation) | Style embedding · Ads text-to-personality | top-3 accuracy: 0.2806 -> 0.3019 | Improved |
| Train style constraints jointly across all modalities rather than per task (ablation) | Style embedding · Ads text-to-gaze | top-3 accuracy: 0.2704 -> 0.2671 | Hurt |
Closest papers by meaning
- Towards Unified Multi-Modal Personalization: Large Vision-Language Models for Generative Recommendation and Beyond (ICLR 2024 · similarity 0.87)
- Aligning Vision Models with Human Aesthetics in Retrieval: Benchmarks and Algorithms (NeurIPS 2024 · similarity 0.87)
- Same Concept, Different Directions: Cross-Modal Feature Heterogeneity in Sparse Autoencoders (NeurIPS 2026 · similarity 0.87)
- Cross the Gap: Exposing the Intra-modal Misalignment in CLIP via Modality Inversion (ICLR 2025 · similarity 0.86)
- Dynamic Adapter with Semantics Disentangling for Cross-lingual Cross-modal Retrieval (AAAI 2025 · similarity 0.86)
- Adaptive Prompt-Based Semantic Embedding with Inspire Potential of Implicit Knowledge for Cross-Modal Retrieval (AAAI 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
