Research map / Image generation / Image restoration
Viewpoint-Agnostic Change Captioning With Cycle Consistency
Hoeseong Kim, Jongseok Kim, Hyungseok Lee, Hyunsung Park, Gunhee Kim
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 cycle consistency module (ablation) | VACC · CLEVR-DC | CIDEr: 70.0 -> 71.7 | Improved |
| add cycle consistency module | DUDA · CLEVR-DC | CIDEr: 56.7 -> 62.0 | Improved |
| add cycle consistency module | M-VAM · CLEVR-DC | CIDEr: 60.1 -> 60.6 | Improved |
| use unchanged features in cycle consistency instead of changed features (ablation) | M-VAM + CC · CLEVR-DC | CIDEr: 60.6 -> 57.1 | Hurt |
| use backbone outputs Xbef/Xaft in cycle consistency instead of xbef/xaft (ablation) | VACC · CLEVR-DC | CIDEr: 71.7 -> 69.0 | Hurt |
| add viewpoint encoding to difference modeling (ablation) | VACC without CC · CLEVR-DC | CIDEr: 69.1 -> 70.0 | Improved |
| add difference modeling to viewpoint encoding (ablation) | VACC without CC · CLEVR-DC | CIDEr: 65.6 -> 70.0 | Improved |
| add cycle consistency module to VP only (ablation) | VP · CLEVR-DC | CIDEr: 65.6 -> 67.7 | Improved |
| add cycle consistency module to Diff only (ablation) | Diff · CLEVR-DC | CIDEr: 69.1 -> 69.7 | Improved |
| add cycle consistency module to VP only (ablation) | VP · CLEVR-DC | SPICE: 16.7 -> 16.6 | No clear effect |
| compare VACC against M-VAM | VACC · CLEVR-Change | CIDEr: 114.9 -> 114.2 | No clear effect |
| compare VACC against M-VAM | VACC · Spot-the-Diff | BLEU-4: 10.1 -> 9.7 | Hurt |
Closest papers by meaning
- Region-aware Difference Distilling with Attribute-guided Contrastive Regularization for Change Captioning (AAAI 2025 · similarity 0.88)
- Revisiting Change Captioning from Self-supervised Global-Part Alignment (AAAI 2025 · similarity 0.87)
- Change3D: Revisiting Change Detection and Captioning from A Video Modeling Perspective (CVPR 2025 · similarity 0.86)
- DECIDER: Difference-aware Contrastive Diffusion Model with Adversarial Perturbations for Image Change Captioning (AAAI 2025 · similarity 0.85)
- VIXEN: Visual Text Comparison Network for Image Difference Captioning (AAAI 2024 · similarity 0.85)
- VisMin: Visual Minimal-Change Understanding (NeurIPS 2024 · similarity 0.85)
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.
