# A One-Shot Reparameterization Method for Reducing the Loss of Tile Pruning on DNNs

arXiv: https://arxiv.org/abs/2207.14545 · HTML: https://mlsanity.com/paper/2207.14545/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 37 references that are in the map.
- First posted on arXiv 29 Jul 2022.

## Builds on (most cited first, 24 of 24)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [Learning both Weights and Connections for Efficient Neural Networks](https://mlsanity.com/paper/1506.02626/) (2015)
- [Mixed Precision Training](https://mlsanity.com/paper/1710.03740/) (2017)
- [CogView: Mastering Text-to-Image Generation via Transformers](https://mlsanity.com/paper/2105.13290/) (2021)
- [Rethinking the Value of Network Pruning](https://mlsanity.com/paper/1810.05270/) (2018)
- [Learning Structured Sparsity in Deep Neural Networks](https://mlsanity.com/paper/1608.03665/) (2016)
- [Importance Estimation for Neural Network Pruning](https://mlsanity.com/paper/1906.10771/) (2019)
- [RepVGG: Making VGG-style ConvNets Great Again](https://mlsanity.com/paper/2101.03697/) (2021)
- [Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech](https://mlsanity.com/paper/2105.06337/) (2021)
- [ThiNet: A Filter Level Pruning Method for Deep Neural Network Compression](https://mlsanity.com/paper/1707.06342/) (2017)
- [ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks](https://mlsanity.com/paper/1908.03930/) (2019)
- [Towards Optimal Structured CNN Pruning via Generative Adversarial Learning](https://mlsanity.com/paper/1903.09291/) (2019)
- [Actionable Models: Unsupervised Offline Reinforcement Learning of Robotic Skills](https://mlsanity.com/paper/2104.07749/) (2021)
- [The Limits and Potentials of Deep Learning for Robotics](https://mlsanity.com/paper/1804.06557/) (2018)
- [Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation](https://mlsanity.com/paper/2106.03153/) (2021)
- [Diverse Branch Block: Building a Convolution as an Inception-like Unit](https://mlsanity.com/paper/2103.13425/) (2021)
- [ResRep: Lossless CNN Pruning via Decoupling Remembering and Forgetting](https://mlsanity.com/paper/2007.03260/) (2020)
- [RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition](https://mlsanity.com/paper/2105.01883/) (2021)
- [Block-Sparse Recurrent Neural Networks](https://mlsanity.com/paper/1711.02782/) (2017)
- [Accelerating Sparse DNN Models without Hardware-Support via Tile-Wise Sparsity](https://mlsanity.com/paper/2008.13006/) (2020)
- [Accelerate CNNs from Three Dimensions: A Comprehensive Pruning Framework](https://mlsanity.com/paper/2010.04879/) (2020)
- [DARB: A Density-Aware Regular-Block Pruning for Deep Neural Networks](https://mlsanity.com/paper/1911.08020/) (2019)

## Abstract (arXiv)

Recently, tile pruning has been widely studied to accelerate the inference of deep neural networks (DNNs). However, we found that the loss due to tile pruning, which can eliminate important elements together with unimportant elements, is large on trained DNNs. In this study, we propose a one-shot reparameterization method, called TileTrans, to reduce the loss of tile pruning. Specifically, we repermute the rows or columns of the weight matrix such that the model architecture can be kept unchanged after reparameterization. This repermutation realizes the reparameterization of the DNN model without any retraining. The proposed reparameterization method combines important elements into the same tile; thus, preserving the important elements after the tile pruning. Furthermore, TileTrans can be seamlessly integrated into existing tile pruning methods because it is a pre-processing method executed before pruning, which is orthogonal to most existing methods. The experimental results demonstrate that our method is essential in reducing the loss of tile pruning on DNNs. Specifically, the accuracy is improved by up to 17% for AlexNet while 5% for ResNet-34, where both models are pre-trained on ImageNet.