HiAP: A Multi-Granular Stochastic Auto-Pruning Framework for Vision Transformers

• URL: http://arxiv.org/abs/2603.12222v1
• Date: Thu, 12 Mar 2026 17:45:38 GMT
• Title: HiAP: A Multi-Granular Stochastic Auto-Pruning Framework for Vision Transformers
• Abstract summary: Vision Transformers require significant computational resources and memory bandwidth, severely limiting their deployment on edge devices.<n>We propose Hierarchical Auto-Pruning (HiAP), a continuous relaxation framework that discovers optimal sub-networks in a single end-to-end training phase.<n>HiAP addresses both the memory-bound overhead of loading large matrices and the compute-bound mathematical operations.
• Score: 3.644142828550762
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision Transformers require significant computational resources and memory bandwidth, severely limiting their deployment on edge devices. While recent structured pruning methods successfully reduce theoretical FLOPs, they typically operate at a single structural granularity and rely on complex, multi-stage pipelines with post-hoc thresholding to satisfy sparsity budgets. In this paper, we propose Hierarchical Auto-Pruning (HiAP), a continuous relaxation framework that discovers optimal sub-networks in a single end-to-end training phase without requiring manual importance heuristics or predefined per-layer sparsity targets. HiAP introduces stochastic Gumbel-Sigmoid gates at multiple granularities: macro-gates to prune entire attention heads and FFN blocks, and micro-gates to selectively prune intra-head dimensions and FFN neurons. By optimizing both levels simultaneously, HiAP addresses both the memory-bound overhead of loading large matrices and the compute-bound mathematical operations. HiAP naturally converges to stable sub-networks using a loss function that incorporates both structural feasibility penalties and analytical FLOPs. Extensive experiments on ImageNet demonstrate that HiAP organically discovers highly efficient architectures, and achieves a competitive accuracy-efficiency Pareto frontier for models like DeiT-Small, matching the performance of sophisticated multi-stage methods while significantly simplifying the deployment pipeline.

Related papers

• Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization [2.3012967702331313]
  Large Language Models (LLMs) have achieved widespread adoption because of their strong reasoning and query-response capabilities.<n> deploying them in embedded and edge computing environments remains challenging because of strict latency, memory, and energy constraints.<n>We propose a hardware-aware, multi-objective structured pruning framework to address these limitations.
  arXiv  Detail & Related papers  (2026-06-08T09:38:13Z)
• Ge$^\ ext{2}$mS-T: Multi-Dimensional Grouping for Ultra-High Energy Efficiency in Spiking Transformer [84.8831358775386]
  Spiking Neural Networks (SNNs) offer superior energy efficiency over Artificial Neural Networks (ANNs)<n>Existing paradigms including ANN-SNN Conversion and Spatial-Temporal Backpropagation (STBP) suffer from inherent limitations.<n>We propose Ge$text2$mS-T, a novel architecture implementing grouped computation across temporal, spatial and network structure dimensions.
  arXiv  Detail & Related papers  (2026-04-10T02:58:46Z)
• Collaborative Multi-Mode Pruning for Vision-Language Models [31.988586741092927]
  Vision-Language Models (VLMs) have advanced rapidly within the unified Transformer architecture, yet their deployment on resource-constrained devices remains challenging.<n>We propose Collaborative Multi-Mode Pruning (CoMP), a novel framework tailored for VLMs by performing joint parameter and token pruning.<n>Our method effectively promotes the performance under high pruning ratios by comparing to the state-of-the-art approaches.
  arXiv  Detail & Related papers  (2026-04-03T10:44:23Z)
• Alternating Gradient Flow Utility: A Unified Metric for Structural Pruning and Dynamic Routing in Deep Networks [52.153950303594684]
  We propose a decoupled kinetic paradigm inspired by Alternating Gradient Flow (AGF)<n>AGF accurately captures the network's structural "kinetic utility"<n>We design a hybrid routing framework that decouples AGF-guided offline structural search from online execution via zero-cost physical priors.
  arXiv  Detail & Related papers  (2026-03-12T18:19:21Z)
• MASPOB: Bandit-Based Prompt Optimization for Multi-Agent Systems with Graph Neural Networks [21.211097851224487]
  We introduce MASPOB (Multi-Agent System Prompt Optimization via Bandits), a novel sample-efficient framework based on bandits.<n>To handle topology-induced coupling, MASPOB integrates Graph Neural Networks (GNNs) to capture structural priors, learning topology-aware representations of prompt semantics.
  arXiv  Detail & Related papers  (2026-03-03T05:59:05Z)
• Discrete-Guided Diffusion for Scalable and Safe Multi-Robot Motion Planning [56.240199425429445]
  Multi-Robot Motion Planning (MPMP) involves generating trajectories for multiple robots operating in a shared continuous workspace.<n>While discrete multi-agent finding (MAPF) methods are broadly adopted due to their scalability, their coarse discretization trajectory quality.<n>This paper tackles limitations of two approaches by introducing discrete MAPF solvers with constrained generative diffusion models.
  arXiv  Detail & Related papers  (2025-08-27T17:59:36Z)
• RPCANet++: Deep Interpretable Robust PCA for Sparse Object Segmentation [51.37553739930992]
  RPCANet++ is a sparse object segmentation framework that fuses the interpretability of RPCA with efficient deep architectures.<n>Our approach unfolds a relaxed RPCA model into a structured network comprising a Background Approximation Module (BAM), an Object Extraction Module (OEM) and an Image Restoration Module (IRM)<n>Experiments on diverse datasets demonstrate that RPCANet++ achieves state-of-the-art performance under various imaging scenarios.
  arXiv  Detail & Related papers  (2025-08-06T08:19:37Z)
• syftr: Pareto-Optimal Generative AI [40.80352098169579]
  syftr is a framework that performs efficient multi-objective search over a broad space of agentic and non-agentic RAG configurations.<n>Syftr finds flows which are on average approximately 9 times cheaper while preserving most of the accuracy of the most accurate flows.
  arXiv  Detail & Related papers  (2025-05-26T17:43:13Z)
• HiLAB: A Hybrid Inverse-Design Framework [0.0]
  HiLAB is a new paradigm for inverse design of nanophotonic structures.<n>It addresses multi-functional device design by generating diverse freeform configurations at reduced simulation costs.
  arXiv  Detail & Related papers  (2025-05-23T05:34:56Z)
• Sculpting Features from Noise: Reward-Guided Hierarchical Diffusion for Task-Optimal Feature Transformation [18.670626228472877]
  DIFFT redefines Feature Transformation as a reward-guided generative task.<n>It produces structured, discrete features, preserving intra-feature dependencies while allowing parallel inter-feature generation.<n>It consistently outperforms state-of-the-art baselines in predictive accuracy and robustness, with significantly lower training and inference times.
  arXiv  Detail & Related papers  (2025-05-21T06:18:42Z)
• ALoRE: Efficient Visual Adaptation via Aggregating Low Rank Experts [71.91042186338163]
  ALoRE is a novel PETL method that reuses the hypercomplex parameterized space constructed by Kronecker product to Aggregate Low Rank Experts.<n>Thanks to the artful design, ALoRE maintains negligible extra parameters and can be effortlessly merged into the frozen backbone.
  arXiv  Detail & Related papers  (2024-12-11T12:31:30Z)
• SHERL: Synthesizing High Accuracy and Efficient Memory for Resource-Limited Transfer Learning [63.93193829913252]
  We propose an innovative METL strategy called SHERL for resource-limited scenarios. In the early route, intermediate outputs are consolidated via an anti-redundancy operation. In the late route, utilizing minimal late pre-trained layers could alleviate the peak demand on memory overhead.
  arXiv  Detail & Related papers  (2024-07-10T10:22:35Z)
• Energy-efficient Task Adaptation for NLP Edge Inference Leveraging Heterogeneous Memory Architectures [68.91874045918112]
  adapter-ALBERT is an efficient model optimization for maximal data reuse across different tasks. We demonstrate the advantage of mapping the model to a heterogeneous on-chip memory architecture by performing simulations on a validated NLP edge accelerator.
  arXiv  Detail & Related papers  (2023-03-25T14:40:59Z)
• Pruning Self-attentions into Convolutional Layers in Single Path [89.55361659622305]
  Vision Transformers (ViTs) have achieved impressive performance over various computer vision tasks. We propose Single-Path Vision Transformer pruning (SPViT) to efficiently and automatically compress the pre-trained ViTs. Our SPViT can trim 52.0% FLOPs for DeiT-B and get an impressive 0.6% top-1 accuracy gain simultaneously.
  arXiv  Detail & Related papers  (2021-11-23T11:35:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.