Action Recognition With Gradient Boundary Convolutional Network

Keywords

Action recognition; Convolutional network; Gradient boundary

Abstract

Deep learning features for video action recognition are usually learned from RGB/gray images, image gradients, and optical flows. The single modality of the input data can describe one characteristic of the human action such as appearance structure or motion information. In this paper, we propose a high efficient gradient boundary convolutional network (ConvNet) to simultaneously learn spatio-temporal feature from the single modality data of gradient boundaries. The gradient boundaries represent both local spacial structure and motion information of action video. The gradient boundaries also have less background noise compared to RGB/gray images and image gradients. Extensive experiments are conducted on two popular and challenging action benchmarks, the UCF101 and the HMDB51 action datasets. The proposed deep gradient boundary feature achieves competitive performances on both benchmarks.

Publication Date

2-20-2018

Publication Title

Proceedings - International Conference on Image Processing, ICIP

Volume

2017-September

Number of Pages

1047-1051

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.1109/ICIP.2017.8296441

Socpus ID

85045315693 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/85045315693

This document is currently not available here.

Share

COinS