Dynamic Multi-Task Learning with Convolutional Neural Network

By Yuchun Fang, Zhengyan Ma, Zhaoxiang Zhang, Xu-Yao Zhang, Xiang Bai. 2017

35 citations recorded by the source corpus.

Multi-task learning and deep convolutional neural network (CNN) have been successfully used in various fields. This paper considers the integration of CNN and multi-task learning in a novel way to further improve the performance of multiple related tasks. Existing multi-task CNN models usually empirically combine different tasks into a group which is then trained jointly with a strong assumption of model commonality. Furthermore, traditional approaches usually only consider small number of tasks with rigid structure, which is not suitable for large-scale applications. In light of this, we propose a dynamic multi-task CNN model to handle these problems. The proposed model directly learns the task relations from data instead of subjective task grouping. Due to its flexible structure, it supports task-wise incremental training, which is useful for efficient training of massive tasks. Specifically, we add a new task transfer connection (TTC) between the layers of each task. The learned TTC is able to reflect the correlation among different tasks guiding the model dynamically adjusting the multiplexing of the information among different tasks. With the help of TTC, multiple related tasks can further boost the whole performance for each other. Experiments demonstrate that the proposed dynamic multi-task CNN model outperforms traditional approaches.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Sung Ju Hwang
2. Jaehong Yoon
3. Eunho Yang
4. Saehoon Kim
5. Eric Eaton
6. Deliang Fan
7. Junshan Zhang
8. Sen Lin

In Computer Vision and Pattern Recognition

1. Sung Ju Hwang
2. Jaehong Yoon
3. Eunho Yang
4. Saehoon Kim
5. Mehrdad Farajtabar
6. Christopher Kanan
7. Piyush Rai
8. Bing Liu

In Radiology, Nuclear Medicine and Imaging

1. Sung Ju Hwang
2. Jaehong Yoon
3. Eunho Yang
4. Mehrdad Farajtabar
5. Piyush Rai
6. Bing Liu
7. Eric Eaton
8. Christopher Kanan

Citations in the corpus

• Scalable and Order-robust Continual Learning with Additive Parameter Decomposition (2019, background): “Recently introduced expansion-based approaches handle this problem by expanding the network capacity as they adapt to new tasks (Rusu et al., 2016; Fang et al., 2017; Yoon et al., 2018; Li et al., 2019).”
• ORACLE: Order Robust Adaptive Continual LEarning (2019, background): “Thus, a dominant approach nowadays is to allow the network to expand its capacity as they adapt to new tasks (Rusu et al., 2016; Fang et al., 2017; Yoon et al., 2018).”