核心要点
- 问题/背景
- 这篇 Nature 论文属于 AI 硬件-软件协同设计方向，目标是让 neural-field reconstruction 这类稀疏信号重建任务摆脱传统数字计算的能耗和并行瓶颈。
- 方法/机制
- 作者使用 resistive-memory computing 构建 co-optimized hardware-software system，在 imaging 和 3D vision 场景中提升神经场重建的能效和并行度。重点不是单个器件指标，而是把存算硬件与重建算法一起设计。
- 结果/证据
- 正式收录价值在于它改变的是 AI 视觉/成像系统的计算路径：用非易失存算硬件承载 neural-field reconstruction，有潜力影响 3D 感知、边缘视觉和低功耗重建工作流。符合仓库对 AI accelerator/hardware 的收录规则。
- 收录价值
- 它不是更高一级，因为影响范围还取决于硬件可制造性、系统规模和更多任务上的复现；但作为 Nature 级硬件-算法协同重建系统，具备突破性。
原始摘要与中文对照
中文对照翻译
使用电阻式存储器实现高效准确的神经场重建。医疗成像、增强现实和虚拟现实以及具身人工智能（AI）等应用都依赖于从稀疏观测中重建复杂信号的能力。这些应用的特点是不完整测量和有限计算资源。传统的数字硬件方法面临以下挑战：显式信号表示需要大量采样和存储，跨冯·诺依曼瓶颈的数据移动主导能耗和延迟，以及基于CMOS（互补金属氧化物半导体）的电路提供有限的并行效率。在本文中，我们提出了一种用于稀疏输入信号重建的软硬件协同优化框架。在软件层面，我们使用神经场通过神经网络隐式表示信号，并通过低秩分解和结构化剪枝进一步压缩。在硬件层面，我们设计了一个基于电阻式存储器的存内计算平台，其特点是包含一个高斯编码器和一个多层感知器处理引擎。高斯编码器利用电阻式存储器固有的随机性实现高效编码，而处理引擎则通过硬件感知量化电路实现精确的权重映射。在一个40纳米256千比特电阻式存储器宏上，该系统在三维计算机断层扫描稀疏重建、新颖视图合成和动态场景新颖视图合成方面，在不影响重建质量的前提下，分别实现了23.5倍、21.0倍和32.3倍的预计能效提升，以及10.8倍、38.8倍和6.2倍的预计并行度提升。这项工作推动了AI驱动的信号重建技术，并为未来高效、稳健的医疗AI和三维视觉应用铺平了道路。
原始摘要
Applications such as medical imaging, augmented and virtual reality, and embodied artificial intelligence (AI) depend on the ability to reconstruct complex signals from sparse observations. These applications are characterized by incomplete measurements and limited computational resources. Traditional approaches to digital hardware face the following challenges: explicit signal representations require heavy sampling and storage, data movement across the von Neumann bottleneck dominates energy and latency, and CMOS (complementary metal–oxide–semiconductor)-based circuits offer limited parallel efficiency. Here we present a software–hardware co-optimization framework for sparse-input signal reconstruction. At the software level, we use neural fields to implicitly represent signals using neural networks, which are further compressed by low-rank decomposition and structured pruning. At the hardware level, we design a resistive-memory-based computing-in-memory platform, featuring a Gaussian encoder and a multi-layer perceptron processing engine. The Gaussian encoder leverages the intrinsic stochasticity of resistive memory for efficient encoding, whereas the processing engine enables precise weight mapping through a hardware-aware quantization circuit. On a 40-nm 256 Kb resistive-memory macro, the system delivers 23.5×, 21.0× and 32.3× gains in projected energy efficiency, together with 10.8×, 38.8× and 6.2× gains in projected parallelism, for three-dimensional computed tomography sparse reconstruction, novel view synthesis and dynamic-scene novel view synthesis, without compromising on reconstruction quality. This work advances AI-driven signal reconstruction technology and paves the way for future efficient and robust medical AI and three-dimensional vision applications.