FLICKER: A Fine-Grained Contribution-Aware Accelerator for Real-Time 3D Gaussian Splatting
🤖 AI Summary
This work addresses the excessive computational overhead of 3D Gaussian splatting on resource-constrained edge devices, where rendering a large number of non-contributing Gaussians impedes real-time performance required for AR/VR applications. To overcome this challenge, the authors propose a hardware-software co-designed, fine-grained contribution-aware acceleration architecture that enables near-pixel-level contribution-driven rendering. Key innovations include adaptive guiding pixels, pixel-rectangle grouping, hierarchical Gaussian testing, and mixed-precision computation. Evaluated against the state-of-the-art accelerator, the proposed design achieves 1.5× speedup, 2.6× higher energy efficiency, and 14% area reduction. Moreover, it outperforms edge GPUs by 19.8× in speed and 26.7× in energy efficiency, demonstrating significant promise for real-time immersive applications on edge platforms.
Application Category
📝 Abstract
Recently, 3D Gaussian Splatting (3DGS) has emerged as a mainstream rendering technique due to its photorealistic quality and low latency. However, processing massive numbers of non-contributing Gaussian points introduces significant computational overhead on resource-limited edge platforms, limiting its deployment in next-generation AR/VR devices. Contribution-based prior skipping alleviates this inefficiency, yet the resulting contribution-testing workload becomes prohibitive for edge execution. In this paper, we present FLICKER, a contribution-aware 3DGS accelerator based on hardware-software co-design. The proposed framework integrates adaptive leader pixels, pixel-rectangle grouping, hierarchical Gaussian testing, and a mixed-precision architecture to enable near pixel-level, contribution-driven rendering with minimal overhead. Experimental results demonstrate up to $1.5\times$ speedup, $2.6\times$ improvement in energy efficiency, and $14%$ area reduction compared with a state-of-the-art accelerator. Compared with a representative edge GPU, FLICKER achieves a $19.8\times$ speedup and $26.7\times$ higher energy efficiency.
Problem
Research questions and friction points this paper is trying to address.
3D Gaussian Splatting
contribution-aware
edge computing
real-time rendering
computational overhead
Innovation
Methods, ideas, or system contributions that make the work stand out.
contribution-aware rendering
hardware-software co-design
adaptive leader pixels
hierarchical Gaussian testing
mixed-precision architecture
🔎 Similar Papers
No similar papers found.
