An initial attempt of combining visual selective attention with deep reinforcement learning

2018-11-11 · Liu Yuezhang, Ruohan Zhang, Dana H. Ballard

Visual attention serves as a means of feature selection mechanism in the perceptual system. Motivated by Broadbent's leaky filter model of selective attention, we evaluate how such mechanism could be implemented and affect the learning process of deep reinforcement learning. We visualize and analyze the feature maps of DQN on a toy problem Catch, and propose an approach to combine visual selective attention with deep reinforcement learning. We experiment with optical flow-based attention and A2C on Atari games. Experiment results show that visual selective attention could lead to improvements in terms of sample efficiency on tested games. An intriguing relation between attention and batch normalization is also discovered.

📄 PDF Abstract BibTeX arXiv:1811.04407

Code (0)

등록된 구현이 없습니다.

Tasks

Atari GamesDeep Reinforcement Learningfeature selectionOptical Flow Estimation reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

Methods 이 논문이 사용한 방법론

Q-Learning Q-Learning is an off-policy temporal difference control algorithm: $$Q\left(S\_{t}, A\_{t}\right) \leftarrow Q\left(S\_{t}, A\_{t}\right) + \alpha\left[R_{t+1} +…

A2C A2C, or Advantage Actor Critic, is a synchronous version of the A3C policy gradient method. As an alternative to the asynchronous…

Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…

Convolution A convolution is a type of matrix operation, consisting of a kernel, a small matrix of weights, that slides over input data performing element-wise multiplication with the…

Batch Normalization 설명 없음

DQN A DQN, or Deep Q-Network, approximates a state-value function in a Q-Learning framework with a neural network. In the Atari…

ACTRESS: Active Retraining for Semi-supervised Visual Grounding

2024-07-03 · Weitai Kang, Mengxue Qu, Yunchao Wei, Yan Yan

Semi-Supervised Visual Grounding (SSVG) is a new challenge for its sparse labeled data with the need for multimodel understanding. A previous study, RefTeacher, makes the first attempt to tackle this task by adopting the…

Binary Classification Visual Grounding

Think Again or Think Longer? Selective Verification for Budget-Aware Reasoning

2026-06-18 · Sajib Acharjee Dip, Dawei Zhou, Liqing Zhang arxiv

Test-time reasoning is increasingly used as a serving-time control knob, but extra reasoning is not uniformly valuable: it can repair failed attempts, waste compute on already-correct answers, or introduce harmful answer…

Reinforcement Learning with Attention that Works: A Self-Supervised Approach

2019-04-06 · Anthony Manchin, Ehsan Abbasnejad, Anton Van Den Hengel

Attention models have had a significant positive impact on deep learning across a range of tasks. However previous attempts at integrating attention with reinforcement learning have failed to produce significant improvem…

Atari Games reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

Quality-aware Selective Fusion Network for V-D-T Salient Object Detection

2024-05-13 · Liuxin Bao, Xiaofei Zhou, Xiankai Lu, Yaoqi Sun 외

Depth images and thermal images contain the spatial geometry information and surface temperature information, which can act as complementary information for the RGB modality. However, the quality of the depth and thermal…

object-detection Object Detection Pseudo Label Salient Object Detection

Play It Back: Iterative Attention for Audio Recognition

2022-10-20 · Alexandros Stergiou, Dima Damen

A key function of auditory cognition is the association of characteristic sounds with their corresponding semantics over time. Humans attempting to discriminate between fine-grained audio categories, often replay the sam…

Audio Classification