mlsanityData to 14 Oct 2026

ViSaRL: Visual Reinforcement Learning Guided by Human Saliency

arXiv 2403.10940 ↗First posted 16 Mar 2024cs.RO cs.LGcited by 12

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Training robots to perform complex control tasks from high-dimensional pixel input using reinforcement learning (RL) is sample-inefficient, because image observations are comprised primarily of task-irrelevant information. By contrast, humans are able to visually attend to task-relevant objects and areas. Based on this insight, we introduce Visual Saliency-Guided Reinforcement Learning (ViSaRL). Using ViSaRL to learn visual representations significantly improves the success rate, sample efficiency, and generalization of an RL agent on diverse tasks including DeepMind Control benchmark, robot manipulation in simulation and on a real robot. We present approaches for incorporating saliency into both CNN and Transformer-based encoders. We show that visual representations learned using ViSaRL are robust to various sources of visual perturbations including perceptual noise and scene variations. ViSaRL nearly doubles success rate on the real-robot tasks compared to the baseline which does not use saliency.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 35 of its 54 references that are in the map. First posted on arXiv 16 Mar 2024.

Builds on

35 of 54 references.

References · 35

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  4. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  5. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,733
  6. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,371
  7. 2017Learning Important Features Through Propagating Activation Differencescited by 1,458
  8. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,240
  9. 2019Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learningcited by 1,204
  10. 2020CURL: Contrastive Unsupervised Representations for Reinforcement Learningcited by 760
  11. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 694
  12. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
23 more
  1. 2020Reinforcement Learning with Augmented Datacited by 468
  2. 2020dm_control: Software and Tasks for Continuous Controlcited by 375
  3. 2022MultiMAE: Multi-modal Multi-task Masked Autoencoderscited by 291
  4. 2023Language-Driven Representation Learning for Roboticscited by 165
  5. 2020Generalization in Reinforcement Learning by Soft Data Augmentationcited by 137
  6. 2024THE COLOSSEUM: A Benchmark for Evaluating Generalization for Robotic Manipulationcited by 133
  7. 2016Human Attention in Visual Question Answering: Do Humans and Deep Networks Look at the Same Regions?cited by 111
  8. 2018Revisiting Video Saliency: A Large-scale Benchmark and a New Modelcited by 67
  9. 2014Deep Gaze I: Boosting Saliency Prediction with Feature Maps Trained on ImageNetcited by 60
  10. 2019Exploratory Not Explanatory: Counterfactual Analysis of Saliency Maps for Deep Reinforcement Learningcited by 60
  11. 2021DeepGaze IIE: Calibrated prediction in and out-of-domain for state-of-the-art saliency modelingcited by 56
  12. 2016DeepGaze II: Reading fixations from deep features trained on object recognitioncited by 55
  13. 2015DeepFix: A Fully Convolutional Neural Network for predicting Human Eye Fixationscited by 48
  14. 2022Look where you look! Saliency-guided Q-networks for generalization in visual Reinforcement Learningcited by 34
  15. 2021CYBORG: Blending Human Saliency Into the Loss Improves Deep Learningcited by 33
  16. 2019Efficient Saliency Maps for Explainable AIcited by 30
  17. 2021Human-Aided Saliency Maps Improve Generalization of Deep Learningcited by 27
  18. 2020Feature Expansive Reward Learning: Rethinking Human Inputcited by 25
  19. 2021Learning Reward Functions from Scale Feedbackcited by 24
  20. 2014Person Re-identification by Saliency Learningcited by 20
  21. 2020Are Gradient-based Saliency Maps Useful in Deep Reinforcement Learning?cited by 9
  22. 2022Abstract-to-Executable Trajectory Translation for One-Shot Task Generalizationcited by 2
  23. 2018Exploring to learn visual saliency: The RL-IAC approachcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 7; 2026: 4.

12 papers.

  1. 2025Look, Focus, Act: Efficient and Robust Robot Learning via Human Gaze and Foveated Vision Transformerscited by 11
  2. 2024Observe Then Act: Asynchronous Active Vision-Action Model for Robotic Manipulationcited by 5
  3. 2025AutoFocus-IL: VLM-based Saliency Maps for Data-Efficient Visual Imitation Learning without Extra Human Annotationscited by 4
  4. 2025GABRIL: Gaze-Based Regularization for Mitigating Causal Confusion in Imitation Learningcited by 3
  5. 2025PANDORA: Diffusion Policy Learning for Dexterous Robotic Piano Playingcited by 2
  6. 2025Gaze on the Prize: Shaping Visual Attention with Return-Guided Contrastive Learningcited by 1
  7. 2025DeGuV: Depth-Guided Visual Reinforcement Learning for Generalization and Interpretability in Manipulationcited by 0
  8. 2025QuickLAP: Quick Language-Action Preference Learning for Semi-Autonomous Agentscited by 0
  9. 2026Neurosymbolic Imitation Learning with Human Guidance: A Privileged Information Approachcited by 0
  10. 2026Fast Human Attention Prediction for Fixation-guided Active Perception in Autonomous Navigationcited by 0
  11. 2026Task-Relevant Representation Decoupling for Visual Reinforcement Learning Generalizationcited by 0
  12. 2026Focus Where It Counts: A Salience-Driven Vision-Language Model for Low Vision Assistancecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025VisRL: Intention-Driven Visual Perception via Reinforced Reasoning
  2. 2022Masked World Models for Visual Control
  3. 2025Pre-trained Visual Representations Generalize Where it Matters in Model-Based Reinforcement Learning
  4. 2025Salience-Invariant Consistent Policy Learning for Generalization in Visual Reinforcement Learning
  5. 2025Object-Centric Action-Enhanced Representations for Robot Visuo-Motor Policy Learning
  6. 2024MENTOR: Mixture-of-Experts Network with Task-Oriented Perturbation for Visual Reinforcement Learning
  7. 2026Patch Policy: Efficient Embodied Control via Dense Visual Representations
  8. 2026Saliency-Guided Representation with Consistency Policy Learning for Visual Unsupervised Reinforcement Learning