Adaptive Depth Networks with Skippable Sub-Paths

• URL: http://arxiv.org/abs/2312.16392v3
• Date: Sat, 28 Sep 2024 09:18:59 GMT
• Title: Adaptive Depth Networks with Skippable Sub-Paths
• Authors: Woochul Kang, Hyungseop Lee,
• Abstract summary: We present a practical approach to adaptive depth networks with minimal training effort. Our approach does not train every target sub-network in an iterative manner. We provide a formal rationale for why the proposed training method can reduce overall prediction errors.
• Score: 1.8416014644193066
• License:
• Abstract: Predictable adaptation of network depths can be an effective way to control inference latency and meet the resource condition of various devices. However, previous adaptive depth networks do not provide general principles and a formal explanation on why and which layers can be skipped, and, hence, their approaches are hard to be generalized and require long and complex training steps. In this paper, we present a practical approach to adaptive depth networks that is applicable to various networks with minimal training effort. In our approach, every hierarchical residual stage is divided into two sub-paths, and they are trained to acquire different properties through a simple self-distillation strategy. While the first sub-path is essential for hierarchical feature learning, the second one is trained to refine the learned features and minimize performance degradation if it is skipped. Unlike prior adaptive networks, our approach does not train every target sub-network in an iterative manner. At test time, however, we can connect these sub-paths in a combinatorial manner to select sub-networks of various accuracy-efficiency trade-offs from a single network. We provide a formal rationale for why the proposed training method can reduce overall prediction errors while minimizing the impact of skipping sub-paths. We demonstrate the generality and effectiveness of our approach with convolutional neural networks and transformers.

Related papers

• ChannelDropBack: Forward-Consistent Stochastic Regularization for Deep Networks [5.00301731167245]
  Existing techniques often require modifying the architecture of the network by adding specialized layers. We present ChannelDropBack, a simple regularization approach that introduces randomness only into the backward information flow. It allows for seamless integration into the training process of any model and layers without the need to change its architecture.
  arXiv  Detail & Related papers  (2024-11-16T21:24:44Z)
• Stimulative Training++: Go Beyond The Performance Limits of Residual Networks [91.5381301894899]
  Residual networks have shown great success and become indispensable in recent deep neural network models. Previous research has suggested that residual networks can be considered as ensembles of shallow networks. We identify a problem that is analogous to social loafing, whereworks within a residual network are prone to exert less effort when working as part of a group compared to working alone.
  arXiv  Detail & Related papers  (2023-05-04T02:38:11Z)
• Slimmable Networks for Contrastive Self-supervised Learning [69.9454691873866]
  Self-supervised learning makes significant progress in pre-training large models, but struggles with small models. We introduce another one-stage solution to obtain pre-trained small models without the need for extra teachers. A slimmable network consists of a full network and several weight-sharing sub-networks, which can be pre-trained once to obtain various networks.
  arXiv  Detail & Related papers  (2022-09-30T15:15:05Z)
• Improving the Trainability of Deep Neural Networks through Layerwise Batch-Entropy Regularization [1.3999481573773072]
  We introduce and evaluate the batch-entropy which quantifies the flow of information through each layer of a neural network. We show that we can train a "vanilla" fully connected network and convolutional neural network with 500 layers by simply adding the batch-entropy regularization term to the loss function.
  arXiv  Detail & Related papers  (2022-08-01T20:31:58Z)
• Unsupervised Domain-adaptive Hash for Networks [81.49184987430333]
  Domain-adaptive hash learning has enjoyed considerable success in the computer vision community. We develop an unsupervised domain-adaptive hash learning method for networks, dubbed UDAH.
  arXiv  Detail & Related papers  (2021-08-20T12:09:38Z)
• Manifold Regularized Dynamic Network Pruning [102.24146031250034]
  This paper proposes a new paradigm that dynamically removes redundant filters by embedding the manifold information of all instances into the space of pruned networks. The effectiveness of the proposed method is verified on several benchmarks, which shows better performance in terms of both accuracy and computational cost.
  arXiv  Detail & Related papers  (2021-03-10T03:59:03Z)
• Learning Neural Network Subspaces [74.44457651546728]
  Recent observations have advanced our understanding of the neural network optimization landscape. With a similar computational cost as training one model, we learn lines, curves, and simplexes of high-accuracy neural networks. With a similar computational cost as training one model, we learn lines, curves, and simplexes of high-accuracy neural networks.
  arXiv  Detail & Related papers  (2021-02-20T23:26:58Z)
• Deep Networks from the Principle of Rate Reduction [32.87280757001462]
  This work attempts to interpret modern deep (convolutional) networks from the principles of rate reduction and (shift) invariant classification. We show that the basic iterative ascent gradient scheme for optimizing the rate reduction of learned features naturally leads to a multi-layer deep network, one iteration per layer. All components of this "white box" network have precise optimization, statistical, and geometric interpretation.
  arXiv  Detail & Related papers  (2020-10-27T06:01:43Z)
• Side-Tuning: A Baseline for Network Adaptation via Additive Side Networks [95.51368472949308]
  Adaptation can be useful in cases when training data is scarce, or when one wishes to encode priors in the network. In this paper, we propose a straightforward alternative: side-tuning.
  arXiv  Detail & Related papers  (2019-12-31T18:52:32Z)
• Differentiable Sparsification for Deep Neural Networks [0.0]
  We propose a fully differentiable sparsification method for deep neural networks. The proposed method can learn both the sparsified structure and weights of a network in an end-to-end manner. To the best of our knowledge, this is the first fully differentiable sparsification method.
  arXiv  Detail & Related papers  (2019-10-08T03:57:04Z)

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.