SketchGraphNet: A Memory-Efficient Hybrid Graph Transformer for Large-Scale Sketch Corpora Recognition

2026-03-08 · Shilong Chen, Mingyuan Li, Zhaoyang Wang, Zhonglin Ye, Haixing Zhao arxiv

This work investigates large-scale sketch recognition from a graph-native perspective, where free-hand sketches are directly modeled as structured graphs rather than raster images or stroke sequences. We propose SketchGraphNet, a hybrid graph neural architecture that integrates local message passing with a memory-efficient global attention mechanism, without relying on auxiliary positional or structural encodings. To support systematic evaluation, we construct SketchGraph, a large-scale benchmark comprising 3.44 million graph-structured sketches across 344 categories, with two variants (A and R) to reflect different noise conditions. Each sketch is represented as a spatiotemporal graph with normalized stroke-order attributes. On SketchGraph-A and SketchGraph-R, SketchGraphNet achieves Top-1 accuracies of 83.62% and 87.61%, respectively, under a unified training configuration. MemEffAttn further reduces peak GPU memory by over 40% and training time by more than 30% compared with Performer-based global attention, while maintaining comparable accuracy.

📄 PDF Abstract BibTeX arXiv:2603.07521

Code (0)

등록된 구현이 없습니다.

H-SGANet: Hybrid Sparse Graph Attention Network for Deformable Medical Image Registration

2024-08-29 · Yufeng Zhou, Wenming Cao

The integration of Convolutional Neural Network (ConvNet) and Transformer has emerged as a strong candidate for image registration, leveraging the strengths of both models and a large parameter space. However, this hybri…

Deformable Medical Image Registration GPU Graph Attention Graph Neural Network+2

Retrieval-Aware Distillation for Transformer-SSM Hybrids

2026-02-11 · Aviv Bick, Eric P. Xing, Albert Gu arxiv

State-space models (SSMs) offer efficient sequence modeling but lag behind Transformers on benchmarks that require in-context retrieval. Prior work links this gap to a small set of attention heads, termed Gather-and-Aggr…

GASim: A Graph-Accelerated Hybrid Framework for Social Simulation

2026-05-08 · Xuan Zhou, Yanhui Sun, Hantao Yao, Allen He 외 arxiv

Large-scale social simulators are essential for studying complex social patterns. Prior work explores hybrid methods to scale up simulations, combining large language models (LLM)-based agents with numerical agent-based …

Expressivity-Efficiency Tradeoffs for Hybrid Sequence Models

2026-03-09 · John Cooper, Ilias Diakonikolas, Mingchen Ma, Frederic Sala arxiv

Hybrid sequence models--combining Transformer and state-space model layers--seek to gain the expressive versatility of attention as well as the computational efficiency of state-space model layers. Despite burgeoning int…

Computational Efficiency

Enabling Homomorphically Encrypted Inference for Large DNN Models

2021-03-30 · Guillermo Lloret-Talavera, Marc Jorda, Harald Servat, Fabian Boemer 외

The proliferation of machine learning services in the last few years has raised data privacy concerns. Homomorphic encryption (HE) enables inference using encrypted data but it incurs 100x-10,000x memory and runtime over…