A multimodal LIBRAS-UFOP Brazilian sign language dataset of minimal pairs using a microsoft Kinect sensor

作者：

Highlights：

• There are limited publicly datasets that provide complete RGB-D and skeleton data.

• We propose a new dataset for dynamic sign language recognition based on minimal pairs concept.

• The dataset provides the complete RGB-D and skeleton data recorded by a Kinect sensor.

• Experimental protocol, baseline method, and initial results are explained.

摘要

•There are limited publicly datasets that provide complete RGB-D and skeleton data.•We propose a new dataset for dynamic sign language recognition based on minimal pairs concept.•The dataset provides the complete RGB-D and skeleton data recorded by a Kinect sensor.•Experimental protocol, baseline method, and initial results are explained.

论文关键词：Sign language dataset,Minimal pairs,Sign language recognition,Dynamic images,RGB-D data,CNN

论文评审过程：Received 15 January 2020, Revised 1 October 2020, Accepted 26 October 2020, Available online 28 October 2020, Version of Record 10 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114179