视觉理解论文系列(三)VL-BERT: PRE-TRAINING OF GENERIC VISUALLINGUISTIC REPRESENTATIONS
背景 这是微软亚研院的工作，将纯文本设定的bert扩展到visual-linguistic场景，从预训练到fine-tune，可以用于多个下游任务。 摘要 作者提出了一个可预训练的用于学习视觉-语言任务通用表征的模型VL-BERT，VL-BERT以transformers为主干，可以同时接受V、L特征作为输入。预训练任务使用了包括visual-language数据集Conceptual Capti
相关文章
- 1. 视觉理解论文系列(四)ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Languag
- 2. Contextual Word Representations and Pretraining
- 3. 【论文阅读】 VL-BERT: Pre-training of generic visual-linguistic representations
- 4. 【cs224n-12】Modeling contexts of use: Contextual Representations and Pretraining. ELMo and BERT.
- 5. NLP系列：Word2Vec原始论文：Efficient Estimation of Word Representations in Vector Space
- 6. 论文阅读：DeepWalk Online Learning of Social Representations
- 7. 论文笔记----DeepWalk: Online Learning of Social Representations
- 8. 论文笔记：DeepWalk: Online Learning of Social Representations
- 9. NLP基础之——Contextual Word Representations and Pretraining
- 10. 论文写作解读：A Simple Framework for Contrastive Learning of Visual Representations
- 更多相关文章...
- • CAP理论是什么？ - NoSQL教程
- • jQuery Mobile 列表视图 - jQuery Mobile 教程
- • 三篇文章了解 TiDB 技术内幕——说存储
- • 三篇文章了解 TiDB 技术内幕 —— 说计算
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. 视觉理解论文系列(四)ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Languag
- 2. Contextual Word Representations and Pretraining
- 3. 【论文阅读】 VL-BERT: Pre-training of generic visual-linguistic representations
- 4. 【cs224n-12】Modeling contexts of use: Contextual Representations and Pretraining. ELMo and BERT.
- 5. NLP系列：Word2Vec原始论文：Efficient Estimation of Word Representations in Vector Space
- 6. 论文阅读：DeepWalk Online Learning of Social Representations
- 7. 论文笔记----DeepWalk: Online Learning of Social Representations
- 8. 论文笔记：DeepWalk: Online Learning of Social Representations
- 9. NLP基础之——Contextual Word Representations and Pretraining
- 10. 论文写作解读：A Simple Framework for Contrastive Learning of Visual Representations