牛津大学是一所英国研究型大学,也是罗素大学集团、英国“G5超级精英大学”,欧洲顶尖大学科英布拉集团、欧洲研究型大学联盟的核心成员。牛津大学培养了众多社会名人,包括了27位英国首相、60位诺贝尔奖得主以及数十位世界各国的皇室成员和政治领袖。2016年9月,泰晤士高等教育发布了2016-2017年度世界大学排名,其中牛津大学排名第一。

VIP内容

题目: A Survey on Contextual Embeddings

摘要:

上下文嵌入,如ELMo和BERT,超越了像Word2Vec这样的全局单词表示,在广泛的自然语言处理任务中取得了突破性的性能。上下文嵌入根据上下文为每个单词分配一个表示,从而捕获不同上下文中单词的用法,并对跨语言传输的知识进行编码。在这项调查中,我们回顾了现有的上下文嵌入模型、跨语言的多语言预训练、上下文嵌入在下游任务中的应用、模型压缩和模型分析。

成为VIP会员查看完整内容
0
33

最新论文

We present LM-Reloc -- a novel approach for visual relocalization based on direct image alignment. In contrast to prior works that tackle the problem with a feature-based formulation, the proposed method does not rely on feature matching and RANSAC. Hence, the method can utilize not only corners but any region of the image with gradients. In particular, we propose a loss formulation inspired by the classical Levenberg-Marquardt algorithm to train LM-Net. The learned features significantly improve the robustness of direct image alignment, especially for relocalization across different conditions. To further improve the robustness of LM-Net against large image baselines, we propose a pose estimation network, CorrPoseNet, which regresses the relative pose to bootstrap the direct image alignment. Evaluations on the CARLA and Oxford RobotCar relocalization tracking benchmark show that our approach delivers more accurate results than previous state-of-the-art methods while being comparable in terms of robustness.

0
0
下载
预览
Top