A Comparative Study of CNN Optimization Methods for Edge AI: Exploring the Role of Early Exits

• URL: http://arxiv.org/abs/2604.14789v1
• Date: Thu, 16 Apr 2026 08:50:49 GMT
• Title: A Comparative Study of CNN Optimization Methods for Edge AI: Exploring the Role of Early Exits
• Abstract summary: This paper presents a unified deployment-oriented comparison of static compression and dynamic early-exit mechanisms.<n>Our results show that static and dynamic techniques offer fundamentally different trade-offs for edge deployment.<n>Their combination proves highly effective, simultaneously reducing inference latency and memory usage with minimal accuracy loss.
• Score: 0.08209843760716957
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Deploying deep neural networks on edge devices requires balancing accuracy, latency, and resource constraints under realistic execution conditions. To fit models within these constraints, two broad strategies have emerged: static compression techniques such as pruning and quantization, which permanently reduce model size, and dynamic approaches such as early-exit mechanisms, which adapt computational cost at runtime. While both families are widely studied in isolation, they are rarely compared under identical conditions on physical hardware. This paper presents a unified deployment-oriented comparison of static compression and dynamic early-exit mechanisms, evaluated on real edge devices using ONNX based inference pipelines. Our results show that static and dynamic techniques offer fundamentally different trade-offs for edge deployment. While pruning and quantization deliver consistent memory footprint reduction, early-exit mechanisms enable input-adaptive computation savings that static methods cannot match. Their combination proves highly effective, simultaneously reducing inference latency and memory usage with minimal accuracy loss, expanding what is achievable at the edge.

Related papers

• When Does Scale-Invariant Optimization Become Unstable? An Exact Schedule Law with Weight Decay [6.625270757109195]
  We show that a single quantity captures all schedule and decay forcing, while norm growth induces an opposing geometric self-quenching effect.<n>This yields a sharp boundary that cleanly separates contraction- and expansion-dominated effective learning rate regimes.
  arXiv  Detail & Related papers  (2026-09-08T17:46:17Z)
• Hybrid Robustness Verification for Spatio-Temporal Neural Networks [19.026662893450425]
  Existing verification methods rely on overly conservative approximations or incur prohibitive computational costs.<n>In practice, adversarial perturbations exhibit structured spatial and temporal correlations, constrained to lower-dimensional, semantically meaningful subspaces.<n>We study robustness verification 3D CNNs processing video inputs, targeting applications in action recognition (UCF-101), autonomous driving (Udacity), medical imaging (MedMNIST)
  arXiv  Detail & Related papers  (2026-06-08T17:06:51Z)
• Low-Rank Compression of Pretrained Models via Randomized Subspace Iteration [2.1016374925364616]
  Low-rank decomposition based on the singular value decomposition (SVD) provides a principled approach for model reduction.<n> Randomized alternatives such as randomized SVD (RSVD) improve efficiency, yet they suffer from poor approximation quality when the singular value spectrum decays slowly.<n>We propose randomized subspace (RSI) as a more effective alternative.
  arXiv  Detail & Related papers  (2026-04-03T02:47:03Z)
• CAMP-HiVe: Cyclic Pair Merging based Efficient DNN Pruning with Hessian-Vector Approximation for Resource-Constrained Systems [3.343542849202802]
  We introduce CAMP-HiVe, a cyclic pair merging-based pruning with Hessian Vector approximation.<n>Our experimental results demonstrate that our proposed method achieves significant reductions in computational requirements.<n>It outperforms the existing state-of-the-art neural pruning methods.
  arXiv  Detail & Related papers  (2025-11-09T07:58:36Z)
• In-memory Training on Analog Devices with Limited Conductance States via Multi-tile Residual Learning [59.091567092071564]
  In-memory training typically requires at least 8-bit conductance states to match digital baselines.<n>Many promising memristive devices such as ReRAM offer only about 4-bit resolution due to fabrication constraints.<n>This paper proposes a emphresidual learning framework that sequentially learns on multiple crossbar tiles to compensate the residual errors.
  arXiv  Detail & Related papers  (2025-10-02T19:44:25Z)
• Integrating Pruning with Quantization for Efficient Deep Neural Networks Compression [0.0]
  pruning and quantization are widely used compression techniques to reduce model size and enhance processing speed.<n>We propose two approaches that integrate similarity-based filter pruning with Adaptive Power-of-Two (APoT) quantization to achieve higher compression efficiency.<n> Experimental results demonstrate that our proposed approaches achieve effective model compression with minimal accuracy degradation.
  arXiv  Detail & Related papers  (2025-09-04T14:17:28Z)
• The Larger the Merrier? Efficient Large AI Model Inference in Wireless Edge Networks [56.37880529653111]
  The demand for large computation model (LAIM) services is driving a paradigm shift from traditional cloud-based inference to edge-based inference for low-latency, privacy-preserving applications.<n>In this paper, we investigate the LAIM-inference scheme, where a pre-trained LAIM is pruned and partitioned into on-device and on-server sub-models for deployment.
  arXiv  Detail & Related papers  (2025-05-14T08:18:55Z)
• Accelerating Linear Recurrent Neural Networks for the Edge with Unstructured Sparsity [39.483346492111515]
  Linear recurrent neural networks enable powerful long-range sequence modeling with constant memory usage and time-per-token during inference.<n>Unstructured sparsity offers a compelling solution, enabling substantial reductions in compute and memory requirements when accelerated by compatible hardware platforms.<n>We find that highly sparse linear RNNs consistently achieve better efficiency-performance trade-offs than dense baselines.
  arXiv  Detail & Related papers  (2025-02-03T13:09:21Z)
• Adaptive Error-Bounded Hierarchical Matrices for Efficient Neural Network Compression [0.0]
  This paper introduces a dynamic, error-bounded hierarchical matrix (H-matrix) compression method tailored for Physics-Informed Neural Networks (PINNs) The proposed approach reduces the computational complexity and memory demands of large-scale physics-based models while preserving the essential properties of the Neural Tangent Kernel (NTK) Empirical results demonstrate that this technique outperforms traditional compression methods, such as Singular Value Decomposition (SVD), pruning, and quantization, by maintaining high accuracy and improving generalization capabilities.
  arXiv  Detail & Related papers  (2024-09-11T05:55:51Z)
• Towards Continual Learning Desiderata via HSIC-Bottleneck Orthogonalization and Equiangular Embedding [55.107555305760954]
  We propose a conceptually simple yet effective method that attributes forgetting to layer-wise parameter overwriting and the resulting decision boundary distortion. Our method achieves competitive accuracy performance, even with absolute superiority of zero exemplar buffer and 1.02x the base model.
  arXiv  Detail & Related papers  (2024-01-17T09:01:29Z)
• Towards a Better Theoretical Understanding of Independent Subnetwork Training [56.24689348875711]
  We take a closer theoretical look at Independent Subnetwork Training (IST) IST is a recently proposed and highly effective technique for solving the aforementioned problems. We identify fundamental differences between IST and alternative approaches, such as distributed methods with compressed communication.
  arXiv  Detail & Related papers  (2023-06-28T18:14:22Z)
• Efficient Micro-Structured Weight Unification and Pruning for Neural Network Compression [56.83861738731913]
  Deep Neural Network (DNN) models are essential for practical applications, especially for resource limited devices. Previous unstructured or structured weight pruning methods can hardly truly accelerate inference. We propose a generalized weight unification framework at a hardware compatible micro-structured level to achieve high amount of compression and acceleration.
  arXiv  Detail & Related papers  (2021-06-15T17:22:59Z)
• ALF: Autoencoder-based Low-rank Filter-sharing for Efficient Convolutional Neural Networks [63.91384986073851]
  We propose the autoencoder-based low-rank filter-sharing technique technique (ALF) ALF shows a reduction of 70% in network parameters, 61% in operations and 41% in execution time, with minimal loss in accuracy.
  arXiv  Detail & Related papers  (2020-07-27T09:01:22Z)

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.