arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.11858cs.CVcs.LG
收藏

MEAL:基于流形嵌入的主动学习

MEAL: Manifold Embedding-based Active Learning

Merantix Labs GmbH(Merantix实验室有限责任公司)

Deepthi Sreenivasaiah, Johannes Otterbach, Thomas Wollmann

摘要:

图像分割是自动驾驶中常见且具有挑战性的任务。训练数据的足够像素级注释的可用性是一个障碍。主动学习通过建议最有希望的样本进行标记来帮助从少量数据中学习。在这项工作中,我们提出了一个新的池为基础的方法主动学习,提出了有前途的补丁提取完整的图像,在每个采集步骤。这个问题是在一个探索-开发的框架,结合嵌入基于均匀流形近似模型的代表性与熵作为不确定性度量模型的信息。我们将我们提出的方法应用于自动驾驶数据集CamVid和Cityscapes,并与最先进的基线进行了定量比较。我们发现,我们的主动学习方法实现了更好的性能相比,以前的方法。

英文摘要:

Image segmentation is a common and challenging task in autonomous driving. Availability of sufficient pixel-level annotations for the training data is a hurdle. Active learning helps learning from small amounts of data by suggesting the most promising samples for labeling. In this work, we propose a new pool-based method for active learning, which proposes promising patches extracted from full image, in each acquisition step. The problem is framed in an exploration-exploitation framework by combining an embedding based on Uniform Manifold Approximation to model representativeness with entropy as uncertainty measure to model informativeness. We applied our proposed method to the autonomous driving datasets CamVid and Cityscapes and performed a quantitative comparison with state-of-the-art baselines. We find that our active learning method achieves better performance compared to previous methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑