Research map / Vision-language and multimodal / Human and embodied modeling
Compare and Reweight: Distinctive Image Captioning Using Similar Images Sets
Jiuniu Wang, Wenjia Xu, Qingzhong Wang, Antoni B. Chan
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 CIDErBtw-weighted XE loss (MLE) | Transformer · MSCOCO test split | CIDErBtw: 80.98 -> 75.35 | Improved |
| add CIDErBtw-weighted XE loss (MLE) | Transformer · MSCOCO test split | CIDEr: 110.13 -> 112.44 | Improved |
| add CIDErBtw-weighted loss and reward (SCST) | Transformer+SCST · MSCOCO test split | CIDEr: 125.13 -> 128.11 | Improved |
| add CIDErBtw-weighted loss and reward (SCST) | Transformer+SCST · MSCOCO test split | CIDErBtw: 86.68 -> 84.70 | Improved |
| add CIDErBtw reward in RL | Transformer+SCST+CIDErBtw · MSCOCO test split | CIDErBtw: 84.70 -> 82.74 | Improved |
| add CIDErBtw reward in RL | Transformer+SCST+CIDErBtw · MSCOCO test split | CIDEr: 128.11 -> 127.78 | Hurt |
| add CIDErBtw reward in RL | Transformer+SCST+CIDErBtw · MSCOCO test split | R@1: 24.46 -> 26.46 | Improved |
| add CIDErBtw-weighted loss (MLE) | FC · MSCOCO test split | CIDEr: 97.90 -> 98.82 | Improved |
| add CIDErBtw-weighted loss (MLE) | UpDown · MSCOCO test split | R@1: 20.10 -> 20.92 | Improved |
| add CIDErBtw weighting to SCST | FC+SCST · MSCOCO test split | CIDErBtw: 90.09 -> 89.41 | Improved |
| train with SCST instead of MLE | Transformer · MSCOCO test split | CIDErBtw: 80.98 -> 86.68 | Hurt |
| train with SCST instead of MLE | Transformer · MSCOCO test split | CIDEr: 110.13 -> 125.13 | Improved |
Closest papers by meaning
- On Distinctive Image Captioning via Comparing and Reweighting (TPAMI 2022 · similarity 0.97)
- Describing Like Humans: On Diversity in Image Captioning (CVPR 2019 · similarity 0.92)
- On Diversity in Image Captioning: Metrics and Methods (TPAMI 2020 · similarity 0.90)
- Group-Based Distinctive Image Captioning with Memory Difference Encoding and Attention (IJCV 2024 · similarity 0.89)
- Describing Differences in Image Sets with Natural Language (CVPR 2024 · similarity 0.86)
- Noise-Aware Image Captioning with Progressively Exploring Mismatched Words (AAAI 2024 · 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.
