Search papers, labs, and topics across Lattice.
This paper introduces SpikeRestormer, an energy-efficient Spiking Neural Network (SNN) designed for All-in-One image restoration (AiOIR) that addresses the high computational costs associated with traditional ANN-based methods. By leveraging a novel degradation-event perception process and integrating mechanisms like Subtractive Degradation Event Attention (SDEA) and Hierarchical Bayesian Skip Masking (HBSM), SpikeRestormer effectively extracts and infers degradation events to enhance image restoration. The results demonstrate that SpikeRestormer achieves competitive performance with ANN methods while significantly reducing energy consumption, setting a new benchmark for SNN-based image restoration techniques.
SpikeRestormer achieves state-of-the-art image restoration performance with significantly lower energy usage, revolutionizing the potential of Spiking Neural Networks in real-time applications.
ANN-based All-in-One image restoration (AiOIR) unifies diverse degradation handling but incurs high computational costs, limiting its real-time deployment. While Spiking Neural Networks (SNNs) offer a low-power alternative, applying them to static images remains challenging. This difficulty arises because explicit event signals are absent, and degradation cues are heavily entangled with scene structures, hindering the learning of reliable restoration-oriented spike events. To address these issues, we propose SpikeRestormer, an energy-efficient SNN for AiOIR that performs event reasoning over internally generated spike cues. Specifically, we propose a degradation-event perception process to extract spike-based degradation events through Subtractive Degradation Event Attention (SDEA). Moreover, we introduce Hierarchical Bayesian Skip Masking (HBSM) and Additive Restoration Event Attention (AREA) processes for event-reliability inference and restoration-event construction, respectively. By integrating these complementary processes, SpikeRestormer formulates restoration as a unified process of degradation-event perception, degradation-event reliability inference, and restoration-event construction, liberating the potential of SNNs for energy-efficient AiOIR. Extensive experiments show that SpikeRestormer delivers competitive performance against ANN-based methods and establishes new state-of-the-art results among SNN-based methods with significantly lower energy consumption.
