Pixel Sentence Representation Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911776214351872 |
|---|---|
| author | Xiao, Chenghao Huang, Zhuoxu Chen, Danlu Hudson, G Thomas Li, Yizhi Duan, Haoran Lin, Chenghua Fu, Jie Han, Jungong Moubayed, Noura Al |
| author_facet | Xiao, Chenghao Huang, Zhuoxu Chen, Danlu Hudson, G Thomas Li, Yizhi Duan, Haoran Lin, Chenghua Fu, Jie Han, Jungong Moubayed, Noura Al |
| contents | Pretrained language models are long known to be subpar in capturing sentence and document-level semantics. Though heavily investigated, transferring perturbation-based methods from unsupervised visual representation learning to NLP remains an unsolved problem. This is largely due to the discreteness of subword units brought by tokenization of language models, limiting small perturbations of inputs to form semantics-preserved positive pairs. In this work, we conceptualize the learning of sentence-level textual semantics as a visual representation learning process. Drawing from cognitive and linguistic sciences, we introduce an unsupervised visual sentence representation learning framework, employing visually-grounded text perturbation methods like typos and word order shuffling, resonating with human cognitive patterns, and enabling perturbation to texts to be perceived as continuous. Our approach is further bolstered by large-scale unsupervised topical alignment training and natural language inference supervision, achieving comparable performance in semantic textual similarity (STS) to existing state-of-the-art NLP methods. Additionally, we unveil our method's inherent zero-shot cross-lingual transferability and a unique leapfrogging pattern across languages during iterative training. To our knowledge, this is the first representation learning method devoid of traditional language models for understanding sentence and document semantics, marking a stride closer to human-like textual comprehension. Our code is available at https://github.com/gowitheflow-1998/Pixel-Linguist |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_08183 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Pixel Sentence Representation Learning Xiao, Chenghao Huang, Zhuoxu Chen, Danlu Hudson, G Thomas Li, Yizhi Duan, Haoran Lin, Chenghua Fu, Jie Han, Jungong Moubayed, Noura Al Computation and Language Computer Vision and Pattern Recognition Pretrained language models are long known to be subpar in capturing sentence and document-level semantics. Though heavily investigated, transferring perturbation-based methods from unsupervised visual representation learning to NLP remains an unsolved problem. This is largely due to the discreteness of subword units brought by tokenization of language models, limiting small perturbations of inputs to form semantics-preserved positive pairs. In this work, we conceptualize the learning of sentence-level textual semantics as a visual representation learning process. Drawing from cognitive and linguistic sciences, we introduce an unsupervised visual sentence representation learning framework, employing visually-grounded text perturbation methods like typos and word order shuffling, resonating with human cognitive patterns, and enabling perturbation to texts to be perceived as continuous. Our approach is further bolstered by large-scale unsupervised topical alignment training and natural language inference supervision, achieving comparable performance in semantic textual similarity (STS) to existing state-of-the-art NLP methods. Additionally, we unveil our method's inherent zero-shot cross-lingual transferability and a unique leapfrogging pattern across languages during iterative training. To our knowledge, this is the first representation learning method devoid of traditional language models for understanding sentence and document semantics, marking a stride closer to human-like textual comprehension. Our code is available at https://github.com/gowitheflow-1998/Pixel-Linguist |
| title | Pixel Sentence Representation Learning |
| topic | Computation and Language Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.08183 |