Search papers, labs, and topics across Lattice.
This paper introduces SpikingNav, a novel framework for robust indoor embodied navigation that leverages spiking neural networks (SNNs) to enhance performance and resilience against visual corruptions. By integrating a Spiking Sensing Encoder (SSE) for feature extraction and a Spiking Policy Network (SPN) for decision-making, SpikingNav demonstrates superior robustness and efficiency compared to traditional artificial neural networks (ANNs). Experimental results reveal a notable increase in navigation success rates, particularly under challenging visual conditions, highlighting the advantages of spike-based computation in resource-constrained environments.
SpikingNav boosts ObjectNav success rates by over 10% under visual corruption, showcasing the power of spiking neural networks in real-world navigation tasks.
Embodied navigation requires an agent to make sequential decisions from egocentric observations in a physical environment. Existing Artificial Neural Network (ANN)-based navigation models have achieved strong performance, yet they often rely on dense computation and may degrade under visual corruptions. Spiking neural networks (SNNs) provide event-driven computation and intrinsic temporal dynamics, which are promising for compact and robust navigation on resource-constrained platforms. However, whether spike-based sensing and policy dynamics can improve robustness in visually rich embodied navigation remains an open problem. This paper proposes SpikingNav, a spiking framework for robust indoor embodied navigation. It contains a Spiking Sensing Encoder (SSE) and a Spiking Policy Network (SPN). The SSE extracts task-conditioned visual features with a spike-based backbone. The SPN maintains a recurrent policy state through membrane integration, thresholding, and spike-triggered reset. In this way, SpikingNav exploits the dynamic properties and spike activations of SNNs to improve navigation performance and robustness. We evaluate SpikingNav on PointNav and ObjectNav under clean observations and visual corruptions. SpikingNav achieves competitive clean performance and stronger robustness with fewer parameters and lower per-step computation than a matched ANN baseline. For instance, SpikingNav improves ObjectNav success from 31.05% to 34.12%, and raises the average success under visual corruptions from 8.45% to 13.71%, demonstrating the benefits of spike-based sensing and policy dynamics. We further validate the deployability of our spike-based sensing method on the Thruster-V2 neuromorphic chip. This physical hardware validation shows that SpikingNav can be instantiated on a real neuromorphic substrate for cyber-physical systems.
