3D Gaussian Splatting (3DGS) has emerged as a mainstream for novel view synthesis, leveraging continuous aggregations of Gaussian functions to model scene geometry. However, 3DGS suffers from substantial memory requirements to store the large amount of Gaussians, hindering its efficiency and practicality. To address this challenge, we introduce GaussianSpa, an optimization-based simplification framework for compact and high-quality 3DGS. Specifically, we formulate the simplification objective as a constrained optimization problem associated with the 3DGS training. Correspondingly, we propose an efficient "optimizing-sparsifying" solution for the formulated problem, alternately solving two independent sub-problems and gradually imposing substantial sparsity onto the Gaussians in the 3DGS training process. We conduct quantitative and qualitative evaluations on various datasets, demonstrating the superiority of GaussianSpa over existing state-of-the-art approaches. Notably, GaussianSpa achieves an average PSNR improvement of 0.9 dB on the real-world Deep Blending dataset with 10× fewer Gaussians compared to the vanilla 3DGS. Our project page is available at https://noodle-lab.github.io/gaussianspa.

Scientific applications generate massive amounts of data, leading to significant storage and I/O bottlenecks. Lossy compression is a crucial technique for reducing data volume in scientific computing, balancing storage efficiency and data fidelity. Such scientific data often consists of multiple fields representing various physical metrics, which exhibit inherent correlations. However, existing compression solutions predominantly rely on local information within a single field, overlooking cross-field correlations that could enhance compression efficiency. In this paper, we introduce a novel compression framework that integrates both local and cross-field information to enhance predictive accuracy. With the help of a carefully designed neural network, our approach effectively captures complex correlations across fields, significantly improving compression ratios. Additionally, our method can rely solely on cross-field prediction to reconstruct fields, achieving extremely high compression ratios under relaxed accuracy constraints. We evaluate our approach on two real-world scientific applications across 85 data fields, demonstrating compression ratio improvements of up to 103.4% for a single field and up to 19.3% overall.

SZ3 (0.07MB, PSNR = 29.2) (c)NeurLZ (0.07MB, PSNR = 39.1)

Mobile devices have become essential enablers for AI applications, particularly in scenarios that require real-time performance. Vision Transformer (ViT) has become a fundamental cornerstone in this regard due to its high accuracy. Recent efforts have been dedicated to developing various transformer architectures that offer im- proved accuracy while reducing the computational requirements. However, existing research primarily focuses on reducing the theoretical computational complexity through methods such as local attention and model pruning, rather than considering realistic performance on mobile hardware. Although these optimizations reduce computational demands, they either introduce additional overheads related to data transformation (e.g., Reshape and Transpose) or irregular computation/data-access patterns. These result in significant overhead on mobile devices due to their limited bandwidth, which even makes the latency worse than vanilla ViT on mobile. In this paper, we present ECP-ViT, a real-time framework that employs the core-periphery principle inspired by the brain functional networks to guide self-attention in ViTs and enable the deployment of ViT models on smartphones. We identify the main bottleneck in transformer structures caused by data transformation and propose a hardware-friendly core-periphery guided self-attention to decrease computation demands. Additionally, we design the system optimizations for intensive data transformation in pruned models. ECP-ViT, with the proposed algorithm-system co-optimizations, achieves a speedup of 4.6× to 26.9× on mobile GPUs across four datasets: STL-10, CIFAR100, TinyImageNet, and ImageNet.