Research map / Vision-language and multimodal / Visual question answering
ADVISE: Symbolism and External Knowledge for Decoding Advertisements
Keren Ye, Adriana Kovashka
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 generic region embedding (ablation) | base triplet · Ads ranking (PSA) | Rank: 4.139 -> 3.444 | Improved |
| add generic region embedding (ablation) | base triplet · Ads ranking (Product) | Rank: 3.110 -> 2.650 | Improved |
| learn regions as symbol anchors instead of generic regions (ablation) | generic region · Ads ranking (PSA) | Rank: 3.444 -> 3.174 | Improved |
| add attention (ablation) | symbol region · Ads ranking (PSA) | Rank: 3.174 -> 3.258 | Hurt |
| add attention (ablation) | symbol region · Ads ranking (Product) | Rank: 2.539 -> 2.488 | Improved |
| add symbol/object constraints (ablation) | symbol region + attention · Ads ranking (PSA) | Rank: 3.258 -> 3.149 | Improved |
| add symbol/object constraints (ablation) | symbol region + attention · Ads ranking (Product) | Rank: 2.488 -> 2.469 | No clear effect |
| add KB symbols additive knowledge (ablation) | symbol/object · Ads ranking (PSA) | Rank: 3.149 -> 3.013 | Improved |
| add KB objects additive knowledge (ablation) | symbol/object · Ads ranking (PSA) | Rank: 3.149 -> 3.108 | Improved |
| add KB symbols additive knowledge (ablation) | symbol/object · Ads ranking (Product) | Rank: 2.469 -> 2.469 | No clear effect |
| hard negative mining (VSE++ vs VSE) | VSE · Ads ranking (PSA) | Rank: 4.155 -> 4.139 | Improved |
| use ADVISE instead of VSE++ | embedding · Ads QA classification (PSA) | top-1 accuracy: 10.03 -> 10.94 | Improved |
Closest papers by meaning
- Automatic Understanding of Image and Video Advertisements (CVPR 2017 · similarity 0.89)
- Interpreting the Rhetoric of Visual Advertisements (TPAMI 2019 · similarity 0.87)
- Rex-Thinker: Grounded Object Referring via Chain-of-Thought Reasoning (ICLR 2026 · similarity 0.81)
- Retrieving Counterfactuals Improves Visual In-Context Learning (CVPR 2026 · similarity 0.80)
- Contextual AD Narration with Interleaved Multimodal Sequence (CVPR 2025 · similarity 0.80)
- A Multimodal Benchmark for Framing of Oil & Gas Advertising and Potential Greenwashing Detection (NeurIPS 2025 · similarity 0.80)
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.
