Research map / Vision-language and multimodal / Human and embodied modeling
On Distinctive Image Captioning via Comparing and Reweighting
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 DCR weighted XE loss | Transformer · MSCOCO test | CIDEr: 110.13 -> 113.98 | Improved |
| add DCR weighted XE loss | Transformer · MSCOCO test | CIDErBtw: 80.98 -> 75.75 | Improved |
| add DCR weighted loss and reward | Transformer+SCST · MSCOCO test | CIDEr: 125.13 -> 128.53 | Improved |
| add DCR weighted loss and reward | Transformer+SCST · MSCOCO test | R@1: 23.38 -> 25.84 | Improved |
| add DCR on weaker model | FC+SCST · MSCOCO test | CIDEr: 104.43 -> 104.61 | No clear effect |
| add CIDErBtw reward (ablation) | Transformer+SCST · MSCOCO test | CIDErBtw: 86.68 -> 82.74 | Improved |
| add long-tailed weights (LTW) (ablation) | Transformer+CIDErBtwReweight · MSCOCO test | CIDEr: 112.44 -> 113.21 | Improved |
| add long-tailed weights (LTW) (ablation) | Transformer+CIDErBtwReweight · MSCOCO test | CIDErBtw: 75.35 -> 75.44 | No clear effect |
| add negative samples (NS) (ablation) | Transformer+LTW · MSCOCO test | CIDEr: 113.21 -> 113.98 | Improved |
| set alpha_w to 0.75 instead of 0 (ablation) | Transformer+SCST+CIDErBtwStrategy · MSCOCO test | CIDEr: 125.38 -> 127.78 | Improved |
| raise alpha_w to 1.25 from 1.00 (ablation) | Transformer+SCST+CIDErBtwStrategy · MSCOCO test | CIDEr: 127.54 -> 125.02 | Hurt |
| add LTW (F_b=5000 vs without LTW) (ablation) | Transformer+DCR · MSCOCO test | CIDEr: 112.82 -> 113.98 | Improved |
Closest papers by meaning
- Compare and Reweight: Distinctive Image Captioning Using Similar Images Sets (ECCV 2020 · similarity 0.97)
- CAPability: A Comprehensive Visual Caption Benchmark for Evaluating Both Correctness and Thoroughness (NeurIPS 2025 · similarity 0.88)
- Noise-Aware Image Captioning with Progressively Exploring Mismatched Words (AAAI 2024 · similarity 0.88)
- A Picture is Worth More Than 77 Text Tokens: Evaluating CLIP-Style Models on Dense Captions (CVPR 2024 · similarity 0.87)
- CaptionQA: Is Your Caption as Useful as the Image Itself? (CVPR 2026 · similarity 0.87)
- Mining Fine-Grained Image-Text Alignment for Zero-Shot Captioning via Text-Only Training (AAAI 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.
