# Where to Look: A Unified Attention Model for Visual Recognition with Reinforcement Learning

arXiv: https://arxiv.org/abs/2111.07169 · HTML: https://mlsanity.com/paper/2111.07169/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 13 of its 25 references that are in the map.
- First posted on arXiv 13 Nov 2021.

## Builds on (most cited first, 13 of 13)

- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Towards Deep Learning Models Resistant to Adversarial Attacks](https://mlsanity.com/paper/1706.06083/) (2017)
- [Feature Pyramid Networks for Object Detection](https://mlsanity.com/paper/1612.03144/) (2016)
- [Show, Attend and Tell: Neural Image Caption Generation with Visual Attention](https://mlsanity.com/paper/1502.03044/) (2015)
- [Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering](https://mlsanity.com/paper/1707.07998/) (2017)
- [Recurrent Models of Visual Attention](https://mlsanity.com/paper/1406.6247/) (2014)
- [Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation](https://mlsanity.com/paper/1604.06057/) (2016)
- [Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks](https://mlsanity.com/paper/1312.6082/) (2013)
- [Fast Minimum-norm Adversarial Attacks through Adaptive Norm Constraints](https://mlsanity.com/paper/2102.12827/) (2021)
- [Video Captioning via Hierarchical Reinforcement Learning](https://mlsanity.com/paper/1711.11135/) (2017)
- [Active Object Localization with Deep Reinforcement Learning](https://mlsanity.com/paper/1511.06015/) (2015)
- [Attentive Contexts for Object Detection](https://mlsanity.com/paper/1603.07415/) (2016)
- [Decorrelated Double Q-learning](https://mlsanity.com/paper/2006.06956/) (2020)

## Built on by (most cited first, 1 of 1)

- [Top-Down Visual Attention from Analysis by Synthesis](https://mlsanity.com/paper/2303.13043/) (2023)

## Abstract (arXiv)

The idea of using the recurrent neural network for visual attention has gained popularity in computer vision community. Although the recurrent attention model (RAM) leverages the glimpses with more large patch size to increasing its scope, it may result in high variance and instability. For example, we need the Gaussian policy with high variance to explore object of interests in a large image, which may cause randomized search and unstable learning. In this paper, we propose to unify the top-down and bottom-up attention together for recurrent visual attention. Our model exploits the image pyramids and Q-learning to select regions of interests in the top-down attention mechanism, which in turn to guide the policy search in the bottom-up approach. In addition, we add another two constraints over the bottom-up recurrent neural networks for better exploration. We train our model in an end-to-end reinforcement learning framework, and evaluate our method on visual classification tasks. The experimental results outperform convolutional neural networks (CNNs) baseline and the bottom-up recurrent attention models on visual classification tasks.