Pix2Struct-large

Reported on 4 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing3 results

Visual Question Answering (VQA)onChartQA
1:1 Accuracy· 2022-10-07
58.6
best: 81.3 (ChartPaLI-5B + PaLM 2-S)
SOTA
Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding arXiv:2210.03347
Visual Question Answering (VQA)onDocVQA test
ANLS· 2022-10-07
0.766
best: 0.9436 (Human)
Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding arXiv:2210.03347
Visual Question Answering (VQA)onInfographicVQA
ANLS· 2022-10-07
40
best: 80.3 (Gemini Ultra (pixel only))
Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding arXiv:2210.03347

Computer Code1 result

Chart Question AnsweringonChartQA
1:1 Accuracy· 2022-10-07
58.6
best: 81.3 (ChartPaLI-5B + PaLM 2-S)
SOTA
Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding arXiv:2210.03347