Chart Question Answering On Plotqa
Metrics
1:1 Accuracy
Results
Performance results of various models on this benchmark
| Paper Title | ||
|---|---|---|
| MatCha4096 + LaMenDa | 92.89 | Synthesize Step-by-Step: Tools Templates and LLMs as Data Generators for Reasoning-Based Chart VQA |
| MatCha | 91.5 | MatCha: Enhancing Visual Language Pretraining with Math Reasoning and Chart Derendering |
| DePlot+FlanPaLM+Codex (PoT Self-Consistency) | 66.6 | DePlot: One-shot visual language reasoning by plot-to-table translation |
| VL-T5-OCR | 66.0 | ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning |
| CRCT | 55.7 | Classification-Regression for Chart Comprehension |
| VisionTapas-OCR | 53.9 | ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning |
0 of 6 row(s) selected.