Learning to Explore using Active Neural SLAM

By Devendra Singh Chaplot, Dhiraj Gandhi, Saurabh Gupta, Abhinav Kumar Gupta, Ruslan Salakhutdinov. 2020

631 citations recorded by the source corpus.

This work presents a modular and hierarchical approach to learn policies for exploring 3D environments, called `Active Neural SLAM'. Our approach leverages the strengths of both classical and learning-based methods, by using analytical path planners with learned SLAM module, and global and local policies. The use of learning provides flexibility with respect to input modalities (in the SLAM module), leverages structural regularities of the world (in global policies), and provides robustness to errors in state estimation (in local policies). Such use of learning within each module retains its benefits, while at the same time, hierarchical decomposition and modular training allow us to sidestep the high sample complexities associated with training end-to-end policies. Our experiments in visually and physically realistic simulated 3D environments demonstrate the effectiveness of our approach over past learning and geometry-based approaches. The proposed model can also be easily transferred to the PointGoal task and was the winning entry of the CVPR 2019 Habitat PointGoal Navigation Challenge.

Subfields: Artificial Intelligence, Control and Systems Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Control and Systems Engineering

Citations in the corpus

A Survey of Embodied AI: From Simulators to Research Tasks (2021, methodology): “Examples of down-stream tasks that make use of visual exploration outputs (i.e. maps) include Image Navigation [26], [73], Point Navigation [11], [44] and Object Navigation [53], [54], [56].”
AllenAct: A Framework for Embodied AI Research (2020, background): “Within these environments, research has progressed towards learning to interact: including visual navigation [12, 13, 14], question answering [15, 16], task completion [17], instruction following [18, 19], language grounding [20, 21], grasping [22, 23], object manipulation [24, 25], future…”
A Survey on Active Simultaneous Localization and Mapping: State of the Art and New Frontiers (2022, background): “[56] used different policies to infer long-term (i.”
The ThreeDWorld Transport Challenge: A Visually Guided Task-and-Motion Planning Benchmark Towards Physically Realistic Embodied AI (2021, neutral): “• Active Exploration: We adapted a baseline from [7].”
Diffusion-based Generation, Optimization, and Planning in 3D Scenes (2023, background): “Among all tasks, visual navigation has been most studied in the vision and robotics community [4, 13, 21, 40, 76, 77, 94, 95].”
Differentiable SLAM-net: Learning Particle SLAM for Visual Navigation (2021, methodology): “Modular approaches include CMP [23], DAN [35] and Neural SLAM [10].”
Neural Topological SLAM for Visual Navigation (2020, methodology): “The results indicate that NTS performs better than both end-to-end RL based baselines [9,47] and methods using metric spatial maps [6, 9].”
Unmanned Aerial Vehicles for Search and Rescue: A Survey (2023, background): “However, the paper [147] stated that this work shows little difference in navigation performance from untrained traditional methods.”
Learning Affordance Landscapes for Interaction Exploration in 3D Environments (2020, background): “The agent is rewarded for maximizing the area covered in its inferred occupancy map [12, 11, 18], the novelty of the states visited [51], pushing the frontier of explored areas [46], and related metrics [47].”
Success Weighted by Completion Time: A Dynamics-Aware Evaluation Criteria for Embodied Navigation (2021, background): “simulators to accomplish tasks such as navigating to a goal point ([4], [9], [17], [20]), or searching for a distant object ([2], [3]), even within previously unseen environments.”
CLIP on Wheels: Zero-Shot Object Navigation as Object Localization and Exploration (2022, background): “…exploration, many navigation tasks are “goal-conditioned,” where the agent needs to navigate to a given position (i.e., point goal [58, 72, 74, 67, 11, 30]), a view of the environment (i.e., image goal [78, 43, 56]), or a specified object category (i.e., object goal [69, 10, 9, 75, 39, 1, 64]).”
Pushing it out of the Way: Interactive Visual Navigation (2021, background): “Most works focus on point navigation (PointNav) [3, 9, 29, 38] or object navigation (ObjectNav) [5, 8, 10, 39].”
Semantic Curiosity for Active Visual Learning (2020, positive): “[9].”
Shaping embodied agent behavior with activity-context priors from egocentric video (2021, background): “Exploration for navigation and interaction Exploration strategies for visual navigation encourage efﬁcient policy learning by rewarding agents for covering area [12, 10, 17], visiting new states [50, 4], expanding the frontier [46], anticipating maps [49], or via intrinsic motivation [41, 6].”
Open-vocabulary Queryable Scene Representations for Real World Planning (2022, background): “A few of these algorithms construct a semantic map of the current region before planning in that region [1], [22]–[24].”
Exploiting Scene-specific Features for Object Goal Navigation (2020, methodology): “For the simpler PointGoal Navigation [1] task, proposed in the “Habitat Challenge 2019”(3), it was possible to observe how a simple model [29] based on LSTM was able to perform better than competitors based on more complex architectures that exploit maps creation [11][8].”
CoWs on Pasture: Baselines and Benchmarks for Language-Driven Zero-Shot Object Navigation (2022, background): “CoW takes inspiration from the semantic mapping line of work [11, 41, 53], and decomposes the navigation task into exploration when the language target is not conﬁdently localized, and target-driven planning otherwise.”