VisualSTS(multi)

• Number of tasks: 2

STS17MultilingualVisualSTS

Semantic Textual Similarity 17 (STS-17) dataset, rendered into images.

Dataset: Pixel-Linguist/rendered-sts17 • License: not specified • Learn more →

Task category	Score	Languages	Domains	Annotations Creators	Sample Creation
image to image (i2i)	cosine_spearman	ara, deu, eng, fra, ita, ... (9)	News, Social, Spoken, Web, Written	human-annotated	rendered

Citation

@article{xiao2024pixel,
  author = {Xiao, Chenghao and Huang, Zhuoxu and Chen, Danlu and Hudson, G Thomas and Li, Yizhi and Duan, Haoran and Lin, Chenghua and Fu, Jie and Han, Jungong and Moubayed, Noura Al},
  journal = {arXiv preprint arXiv:2402.08183},
  title = {Pixel Sentence Representation Learning},
  year = {2024},
}

STSBenchmarkMultilingualVisualSTS

Semantic Textual Similarity Benchmark (STSbenchmark) dataset, translated into target languages using DeepL API,then rendered into images.built upon multi-sts created by Philip May

Dataset: Pixel-Linguist/rendered-stsb • License: not specified • Learn more →

Task category	Score	Languages	Domains	Annotations Creators	Sample Creation
image to image (i2i)	cosine_spearman	cmn, deu, eng, fra, ita, ... (10)	News, Social, Spoken, Web, Written	human-annotated	rendered

Citation

@article{xiao2024pixel,
  author = {Xiao, Chenghao and Huang, Zhuoxu and Chen, Danlu and Hudson, G Thomas and Li, Yizhi and Duan, Haoran and Lin, Chenghua and Fu, Jie and Han, Jungong and Moubayed, Noura Al},
  journal = {arXiv preprint arXiv:2402.08183},
  title = {Pixel Sentence Representation Learning},
  year = {2024},
}