TDAF: Top-Down Attention Framework for Vision Tasks

12/14/2020
by   Bo Pang, et al.
14

Human attention mechanisms often work in a top-down manner, yet it is not well explored in vision research. Here, we propose the Top-Down Attention Framework (TDAF) to capture top-down attentions, which can be easily adopted in most existing models. The designed Recursive Dual-Directional Nested Structure in it forms two sets of orthogonal paths, recursive and structural ones, where bottom-up spatial features and top-down attention features are extracted respectively. Such spatial and attention features are nested deeply, therefore, the proposed framework works in a mixed top-down and bottom-up manner. Empirical evidence shows that our TDAF can capture effective stratified attention information and boost performance. ResNet with TDAF achieves 2.0 improvements on ImageNet. For object detection, the performance is improved by 2.7 for action recognition, the 3D-ResNet adopting TDAF achieves improvements of 1.7

READ FULL TEXT

page 1

page 2

page 4

research
06/20/2019

Human vs Machine Attention in Neural Networks: A Comparative Study

Recent years have witnessed a surge in the popularity of attention mecha...
research
12/10/2021

Global Attention Mechanism: Retain Information to Enhance Channel-Spatial Interactions

A variety of attention mechanisms have been studied to improve the perfo...
research
07/13/2017

Leveraging the Path Signature for Skeleton-based Human Action Recognition

Human action recognition in videos is one of the most challenging tasks ...
research
09/11/2018

Temporal-Spatial Mapping for Action Recognition

Deep learning models have enjoyed great success for image related comput...
research
10/27/2018

A^2-Nets: Double Attention Networks

Learning to capture long-range relations is fundamental to image/video r...
research
07/16/2020

InfoFocus: 3D Object Detection for Autonomous Driving with Dynamic Information Modeling

Real-time 3D object detection is crucial for autonomous cars. Achieving ...

Please sign up or login with your details

Forgot password? Click here to reset