Research map / Vision-language and multimodal / Cross-modal learning
Object-Aware Query Perturbation for Cross-Modal Image-Text Retrieval
Naoya Sogi, Takashi Shibata, Makoto Terao
Read the paperCheck your idea against 137,217 papers
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 |
|---|---|---|---|
| add object-aware query perturbation to BLIP2 cross-attention | BLIP2 · Flickr-30K T2I small objects | R@1: 81.33 -> 84.00 | Improved |
| add Q-Perturbation to BLIP2 | BLIP2 · Flickr-30K T2I overall | R@1: 89.76 -> 89.86 | No clear effect |
| add Q-Perturbation to BLIP2 | BLIP2 · COCO 5K T2I overall | R@1: 68.25 -> 68.35 | No clear effect |
| add Q-Perturbation to COCA | COCA · Flickr-30K T2I | R@1: 77.64 -> 77.68 | No clear effect |
| add Q-Perturbation to InternVL-G | InternVL-G · Flickr-30K I2T | R@1: 94.90 -> 95.40 | Improved |
| use automatically detected boxes (CO-DETR/CO-DINO) instead of annotated boxes (ablation) | Q-Perturbation · Flickr-30K T2I | R@1: 89.82 -> 89.86 | No clear effect |
| enhance queries in K-subspace instead of projecting onto its orthogonal complement (ablation) | Q-Perturbation · Flickr-30K T2I small objects | R@1 (~10%): 77.33 -> 84.00 | Improved |
| perturb queries with K-subspace instead of adding separate object-embedding similarity scores (OEs) (ablation) | BLIP2 · Flickr-30K T2I | R@1: 78.48 -> 89.82 | Improved |
| vary scale, weight function and subspace dimension hyperparameters (ablation) | Q-Perturbation · Flickr-30K T2I | mR@1: 89.44 to 90.00 | No clear effect |
| add Q-Perturbation | BLIP2 inference · visual feature extraction time on GTX 1080Ti | milliseconds: 321.0 -> 413.0 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- BLIP2 — BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation (ICML 2022) · better on 9 of 9 shared results
- COCA — COCA: COllaborative CAusal Regularization for Audio-Visual Question Answering (AAAI 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Adaptive Cross-Modal Prototypes for Cross-Domain Visual-Language Retrieval (CVPR 2021 · similarity 0.89)
- WikiDO: A New Benchmark Evaluating Cross-Modal Retrieval for Vision-Language Models (NeurIPS 2024 · similarity 0.89)
- Cross-Modal and Uni-Modal Soft-Label Alignment for Image-Text Retrieval (AAAI 2024 · similarity 0.88)
- Visual-Text Cross Alignment: Refining the Similarity Score in Vision-Language Models (ICML 2024 · similarity 0.87)
- Dynamic Adapter with Semantics Disentangling for Cross-lingual Cross-modal Retrieval (AAAI 2025 · similarity 0.87)
- Retrieval Across Any Domains via Large-scale Pre-trained Model (ICML 2024 · similarity 0.87)
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.
