RIPE-MambaSpike: Resolution-Independent Spiking-State-Space Interfaces for Parameter-Efficient Event-Based Vision

2026-09-26

AI Summary

This study addresses the parameter redundancy in spiking hybrid models caused by resolution-dependent projections by proposing RIPE-MambaSpike. This method introduces a pioneering input-resolution-independent interface design that replaces conventional projections with a hierarchical multi-resolution bridge of fixed channel width, thereby maintaining a constant parameter count. Furthermore, it incorporates a reparameterized spiking phase, temporally decoupled modulation, and a dual-stream membrane potential attention mechanism constrained by dynamic convex hulls. Experimental results demonstrate that the proposed model achieves Pareto optimality across multiple benchmarks, attaining a top accuracy of 45.7% on DailyDVS-200. Notably, it reduces the parameter count by 3 to 15 times compared to dense artificial neural networks (ANNs), significantly lowering computational complexity while preserving high accuracy.

Abstract

Spiking-Mamba hybrids reach strong accuracy on event-based vision, but existing designs often require tens of millions of parameters. Much of that cost comes from how the spiking front-end is connected to the state-space backbone rather than from the hybrid architecture itself. In a representative model, a single resolution-dependent projection accounts for 33.55M of 36.25M parameters. To that end, we introduce RIPE-MambaSpike (Resolution-Independent, Parameter-Efficient), which replaces that projection with a hierarchical multi-resolution bridge of fixed channel width. Its deployed footprint is 0.870M parameters, constant at fixed time steps and widths across a 43x range of input areas. Reparameterized spiking stages, temporal decoupled modulation, and a dynamic convex-hull-bounded dual-stream membrane-potential attention preserve accuracy under this compact design. Result-wise, RIPE-MambaSpike is pareto-optimal on CIFAR10-DVS, N-Caltech101, and DailyDVS-200. Notably, on the 200-class DailyDVS-200, a scaled 8.04M configuration achieves 45.7% top-1 accuracy, the best reported spiking result on that benchmark, and outperforms prior spiking methods with 3.0-15.1x fewer parameters than dense ANNs. Overall, our findings demonstrate that competitive event-based recognition does not require resolution-dependent parameter growth. Code is available at https://github.com/MuhiminOsim/RIPE-MambaSpike.

Problem

• Event-based Vision
• Parameter Efficiency
• Spiking Neural Networks
• State-Space Models
• Resolution-Independent

Innovation

• Parameter-Efficient
• Resolution-Independent
• Spiking-Mamba Hybrid
• Event-Based Vision
• Membrane-Potential Attention