paper-with-me

Papers

VER: Scaling On-Policy RL Leads to the Emergence of Navigation in Embodied Rearrangement

2022-10-11 · Erik Wijmans, Irfan Essa, Dhruv Batra

We present Variable Experience Rollout (VER), a technique for efficiently scaling batched on-policy reinforcement learning in heterogenous environments (where different environments take vastly different times to generate rollouts) to many GPUs residing on, potentially, many machines. VER combines the strengths of and blurs the line between synchronous and asynchronous on-policy RL methods (SyncOnRL and AsyncOnRL, respectively). VER learns from on-policy experience (like SyncOnRL) and has no synchronization points (like AsyncOnRL). VER leads to significant and consistent speed-ups across a broad range of embodied navigation and mobile manipulation tasks in photorealistic 3D simulation environments. Specifically, for PointGoal navigation and ObjectGoal navigation in Habitat 1.0, VER is 60-100% faster (1.6-2x speedup) than DD-PPO, the current state of art distributed SyncOnRL, with similar sample efficiency. For mobile manipulation tasks (open fridge/cabinet, pick/place objects) in Habitat 2.0 VER is 150% faster (2.5x speedup) on 1 GPU and 170% faster (2.7x speedup) on 8 GPUs than DD-PPO. Compared to SampleFactory (the current state-of-the-art AsyncOnRL), VER matches its speed on 1 GPU, and is 70% faster (1.7x speedup) on 8 GPUs with better sample efficiency. We leverage these speed-ups to train chained skills for GeometricGoal rearrangement tasks in the Home Assistant Benchmark (HAB). We find a surprising emergence of navigation in skills that do not ostensible require any navigation. Specifically, the Pick skill involves a robot picking an object from a table. During training the robot was always spawned close to the table and never needed to navigate. However, we find that if base movement is part of the action space, the robot learns to navigate then pick an object in new environments with 50% success, demonstrating surprisingly high out-of-distribution generalization.

📄 PDF Abstract BibTeX arXiv:2210.05064

Code (1)

facebookresearch/habitat-lab 공식 구현 pytorch

Tasks

GPUNavigateObjectGoal NavigationOut-of-Distribution GeneralizationPointGoal Navigation

Methods 이 논문이 사용한 방법론

BASE 설명 없음
DD-PPO Decentralized Distributed Proximal Policy Optimization (DD-PPO) is a method for distributed reinforcement learning in resource-intensive simulated environments. DD-PPO is…
SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…

Similar Papers 제목 키워드 기반

Emergence of Abstract State Representations in Embodied Sequence Modeling

2023-11-03 · Tian Yun, Zilai Zeng, Kunal Handa, Ashish V. Thapliyal 외

Decision making via sequence modeling aims to mimic the success of language models, where actions taken by an embodied agent are modeled as tokens to predict. Despite their promising performance, it remains unclear if em…

Decision Making

What if Eye...? Computationally Recreating Vision Evolution

2025-01-25 · Kushagra Tiwary, Aaron Young, Zaid Tasneem, Tzofi Klinghoffer 외

Vision systems in nature show remarkable diversity, from simple light-sensitive patches to complex camera eyes with lenses. While natural selection has produced these eyes through countless mutations over millions of yea…

Evolutionary Algorithms

NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Models

2023-05-26 · Gengze Zhou, Yicong Hong, Qi Wu

Trained with an unprecedented scale of data, large language models (LLMs) like ChatGPT and GPT-4 exhibit the emergence of significant reasoning abilities from model scaling. Such a trend underscored the potential of trai…

Instruction FollowingVision and Language NavigationVisual Navigation

Towards Versatile Embodied Navigation

2022-10-30 · Hanqing Wang, Wei Liang, Luc van Gool, Wenguan Wang

With the emergence of varied visual navigation tasks (e.g, image-/object-/audio-goal and vision-language navigation) that specify the target in different ways, the community has made appealing advances in training specia…

Decision MakingVision-Language NavigationVisual Navigation

$NavA^3$: Understanding Any Instruction, Navigating Anywhere, Finding Anything

2025-08-06 · Lingfeng Zhang, Xiaoshuai Hao, Yingbo Tang, Haoxiang Fu 외 arxiv

Embodied navigation is a fundamental capability of embodied intelligence, enabling robots to move and interact within physical environments. However, existing navigation tasks primarily focus on predefined object navigat…

Instruction FollowingObject Localization