Joint Embedding with Multi-Task Learning for Multi-Label Zero-Shot Action Recognition

2018 ◽  
Author(s):  
Rongqiao An ◽  
Zhenjiang Miao ◽  
Qingyu Li
Keyword(s):  
Action Recognition ◽  
Task Learning ◽  
Joint Embedding

scholarly journals Fast and Accurate Driver Action Recognition with Multi-Task Learning of Driver Pose and Action

2021 ◽  
Vol 36 (2) ◽  
pp. A-K93_1-10
Author(s):  
Kenta Nishiyuki ◽  
Tadashi Hyuga ◽  
Hiroshi Tasaki ◽  
Koichi Kinoshita ◽  
Yuki Hasegawa ◽  
...  
Keyword(s):  
Action Recognition ◽  
Task Learning

scholarly journals Multi-Task Learning of Generalizable Representations for Video Action Recognition

2020 ◽  
Author(s):  
Zhiyu Yao ◽  
Yunbo Wang ◽  
Mingsheng Long ◽  
Jianmin Wang ◽  
Philip S. Yu ◽  
...  
Keyword(s):  
Action Recognition ◽  
Task Learning

scholarly journals Optimization of Action Recognition Model Based on Multi-Task Learning and Boundary Gradient

Electronics ◽  
2021 ◽  
Vol 10 (19) ◽  
pp. 2380
Author(s):  
Yiming Xu ◽  
Fangjie Zhou ◽  
Li Wang ◽  
Wei Peng ◽  
Kai Zhang
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Action Recognition ◽  
Experimental Result ◽  
Detection Accuracy ◽  
Convolutional Network ◽  
Action Detection ◽  
Feature Extraction Method ◽  
Task Learning ◽  
Temporal Action

Recently, people’s demand for action recognition has extended from the initial high classification accuracy to the high accuracy of the temporal action detection. It is challenging to meet the two requirements simultaneously. The key to behavior recognition lies in the quantity and quality of the extracted features. In this paper, a two-stream convolutional network is used. A three-dimensional convolutional neural network (3D-CNN) is used to extract spatiotemporal features from the consecutive frames. A two-dimensional convolutional neural network (2D-CNN) is used to extract spatial features from the key-frames. The integration of the two networks is excellent for improving the model’s accuracy and can complete the task of distinguishing the start–stop frame. In this paper, a multi-scale feature extraction method is presented to extract more abundant feature information. At the same time, a multi-task learning model is introduced. It can further improve the accuracy of classification via sharing the data between multiple tasks. The experimental result shows that the accuracy of the modified model is improved by 10%. Meanwhile, we propose the confidence gradient, which can optimize the distinguishing method of the start–stop frame to improve the temporal action detection accuracy. The experimental result shows that the accuracy has been enhanced by 11%.

Multi-Domain and Multi-Task Learning for Human Action Recognition

2019 ◽  
Vol 28 (2) ◽  
pp. 853-867 ◽  
Author(s):  
An-An Liu ◽  
Ning Xu ◽  
Wei-Zhi Nie ◽  
Yu-Ting Su ◽  
Yong-Dong Zhang
Keyword(s):  
Action Recognition ◽  
Human Action Recognition ◽  
Human Action ◽  
Task Learning

Single/multi-view human action recognition via regularized multi-task learning

2015 ◽  
Vol 151 ◽  
pp. 544-553 ◽  
Author(s):  
An-An Liu ◽  
Ning Xu ◽  
Yu-Ting Su ◽  
Hong Lin ◽  
Tong Hao ◽  
...  
Keyword(s):  
Action Recognition ◽  
Human Action Recognition ◽  
Human Action ◽  
Task Learning

scholarly journals A 3D-CNN and LSTM Based Multi-Task Learning Architecture for Action Recognition

IEEE Access ◽  
2019 ◽  
Vol 7 ◽  
pp. 40757-40770 ◽  
Author(s):  
Xi Ouyang ◽  
Shuangjie Xu ◽  
Chaoyun Zhang ◽  
Pan Zhou ◽  
Yang Yang ◽  
...  
Keyword(s):  
Action Recognition ◽  
Task Learning ◽  
3D Cnn

Partwise bag‐of‐words‐based multi‐task learning for human action recognition

2013 ◽  
Vol 49 (13) ◽  
pp. 803-805 ◽  
Author(s):  
An‐An Liu ◽  
Yuting Su ◽  
Zan Gao ◽  
Tong Hao ◽  
Zhao‐Xuan Yang ◽  
...  
Keyword(s):  
Action Recognition ◽  
Human Action Recognition ◽  
Human Action ◽  
Bag Of Words ◽  
Task Learning
Sign in / Sign up

Export Citation Format

Share Document