LEVIR-CC
Models compare aligned aerial images from Texas and describe what changed. Five annotators wrote reference captions for each of the 1,929 test pairs.
The ChangePixel team ran Qwen2-VL, InternVL3, and Pixtral on the test split without training them on LEVIR-CC.
- Made by
- Chenyang Liu, Rui Zhao, Hao Chen, Zhengxia Zou, and Zhenwei Shi
- Released
- April 2022
- Kind
- Detection and captioning
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Subject
- Earth and environmental sciences
- Language
- English
- In and out
- Images, Maps and geodata → Text
- Interaction
- Single turn
- Items from
- Real work and use, Written by crowdworkers
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Vision
RSICCformer tops the leaderboard with 62.8
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The paper reports several caption metrics, so this headline is its BLEU-4 column. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 1 RSICCformer 62.8
- 2 DUDA 57.8
- 3 Capt-Dual-Att 57.5
- 4 MCCFormers-S 56.7
- 5 MCCFormers-D 56.4
- 6 Baseline 56.4
- 7 Capt-Att 53.1
- 8 Capt-Rep-Diff 47.4
Source: The LEVIR-CC paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | RSICCformer | 62.8 | Official leaderboard | ||
| 2 | DUDA | 57.8 | Official leaderboard | ||
| 3 | Capt-Dual-Att | 57.5 | Official leaderboard | ||
| 4 | MCCFormers-S | 56.7 | Official leaderboard | ||
| 5 | MCCFormers-D | 56.4 | Official leaderboard | ||
| 6 | Baseline | 56.4 | Official leaderboard | ||
| 7 | Capt-Att | 53.1 | Official leaderboard | ||
| 8 | Capt-Rep-Diff | 47.4 | Official leaderboard |
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Vision
- MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy.
Qwen 3.6 Plus 86%
- MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
- CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
- MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy.
Seed2.1 Pro 90.7%
- MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy.
Qwen 3.8 Max 97.7%
- MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy.
Qwen 3.8 Max 85.9%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the benchmark’s paper (8 entries).
- Pass Models scored 8 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on levir.buaa.edu.cn: 8 entries at the last read, none a model we track.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.