Research map / Vision-language and multimodal / Human and embodied modeling
Multi-Level Multimodal Common Semantic Space for Image-Phrase Grounding
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use the proposed multilevel attention and common-space method rather than the strongest listed prior baseline | Phrase grounding · VisualGenome test | pointing game accuracy: 30.03 -> 55.16 | Helped |
| Use the proposed multilevel attention and common-space method rather than the strongest listed prior baseline | Phrase grounding · Flickr30k test | pointing game accuracy: 50.10 -> 69.19 | Helped |
| Use the proposed multilevel attention and common-space method rather than the strongest listed prior baseline | Phrase grounding · ReferIt test | pointing game accuracy: 39.98 -> 62.76 | Helped |
| Add multilevel visual features with learned level selection instead of fixed middle/last layers (ablation) | Full model · Flickr30k | pointing game accuracy: 62.67 -> 67.73 | Helped |
| Add nonlinear mapping for both text and visual features instead of linear mappings (ablation) | Common-space model · Flickr30k | pointing game accuracy: 54.75 -> 67.73 | Helped |
| Add nonlinear visual mapping rather than using a linear visual mapping (ablation) | Common-space model with ELMo and fixed-level features · Flickr30k | pointing game accuracy: 56.92 -> 62.67 | Helped |
| Add nonlinear text mapping rather than using a linear text mapping (ablation) | Common-space model with ELMo and fixed-level features · Flickr30k | pointing game accuracy: 58.40 -> 62.67 | Helped |
| Use pretrained contextualized ELMo embeddings instead of a trainable BiLSTM (ablation) | Phrase grounding model · Flickr30k | pointing game accuracy: 61.13 -> 67.73 | Helped |
| Avoid softmax on attention heatmaps (ablation) | Phrase grounding attention · Flickr30k | pointing game accuracy: 47.20 -> 62.67 | Helped |
Closest papers by meaning
- Your Large Vision-Language Model Only Needs A Few Attention Heads For Visual Grounding (CVPR 2025 · similarity 0.91)
- GroundingME: Exposing the Visual Grounding Gap in MLLMs through Multi-Dimensional Evaluation (CVPR 2026 · similarity 0.89)
- Improved Visual Grounding through Self-Consistent Explanations (CVPR 2024 · similarity 0.89)
- Grounding Multimodal Large Language Models to the World (ICLR 2024 · similarity 0.89)
- Empower Words: DualGround for Structured Phrase and Sentence-Level Temporal Grounding (NeurIPS 2025 · similarity 0.89)
- GeM-VG: Towards Generalized Multi-image Visual Grounding with Multimodal Large Language Models (AAAI 2026 · similarity 0.89)
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.