# Deep Network Pruning: A Comparative Study on CNNs in Face Recognition

arXiv: https://arxiv.org/abs/2405.18302 · HTML: https://mlsanity.com/paper/2405.18302/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 36 references that are in the map.
- First posted on arXiv 28 May 2024.

## Builds on (most cited first, 23 of 23)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices](https://mlsanity.com/paper/1707.01083/) (2017)
- [SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size](https://mlsanity.com/paper/1602.07360/) (2016)
- [VGGFace2: A dataset for recognising faces across pose and age](https://mlsanity.com/paper/1710.08092/) (2017)
- [MS-Celeb-1M: A Dataset and Benchmark for Large-Scale Face Recognition](https://mlsanity.com/paper/1607.08221/) (2016)
- [Do Better ImageNet Models Transfer Better?](https://mlsanity.com/paper/1805.08974/) (2018)
- [Importance Estimation for Neural Network Pruning](https://mlsanity.com/paper/1906.10771/) (2019)
- [MobileFaceNets: Efficient CNNs for Accurate Real-Time Face Verification on Mobile Devices](https://mlsanity.com/paper/1804.07573/) (2018)
- [RetinaFace: Single-stage Dense Face Localisation in the Wild](https://mlsanity.com/paper/1905.00641/) (2019)
- [MixConv: Mixed Depthwise Convolutional Kernels](https://mlsanity.com/paper/1907.09595/) (2019)
- [Discrimination-aware Network Pruning for Deep Model Compression](https://mlsanity.com/paper/2001.01050/) (2020)
- [MixFaceNets: Extremely Efficient Face Recognition Networks](https://mlsanity.com/paper/2107.13046/) (2021)
- [PocketNet: Extreme Lightweight Face Recognition Network using Neural Architecture Search and Multi-Step Knowledge Distillation](https://mlsanity.com/paper/2108.10710/) (2021)
- [VarGFaceNet: An Efficient Variable Group Convolutional Neural Network for Lightweight Face Recognition](https://mlsanity.com/paper/1910.04985/) (2019)
- [MobiFace: A Lightweight Deep Learning Face Recognition on Mobile Devices](https://mlsanity.com/paper/1811.11080/) (2018)
- [SqueezeFacePoseNet: Lightweight Face Verification Across Different Poses for Mobile Platforms](https://mlsanity.com/paper/2007.08566/) (2020)
- [VarGNet: Variable Group Convolutional Neural Network for Efficient Embedded Computing](https://mlsanity.com/paper/1907.05653/) (2019)
- [Model Compression Techniques in Biometrics Applications: A Survey](https://mlsanity.com/paper/2401.10139/) (2024)
- [An Explainable Model-Agnostic Algorithm for CNN-based Biometrics Verification](https://mlsanity.com/paper/2307.13428/) (2023)
- [SqueezerFaceNet: Reducing a Small Face Recognition CNN Even More Via Filter Pruning](https://mlsanity.com/paper/2307.10697/) (2023)
- [Exploring the Interchangeability of CNN Embedding Spaces](https://mlsanity.com/paper/2010.02323/) (2020)

## Built on by (most cited first, 3 of 3)

- [Combined CNN and ViT features off-the-shelf: Another astounding baseline for recognition](https://mlsanity.com/paper/2407.19472/) (2024)
- [PreFIQs: Face Image Quality Is What Survives Pruning](https://mlsanity.com/paper/2605.13396/) (2026)
- [Benchmarking CLIP for Zero-Shot Face and Periocular Gender Estimation](https://mlsanity.com/paper/2610.06102/) (2026)

## Abstract (arXiv)

The widespread use of mobile devices for all kinds of transactions makes necessary reliable and real-time identity authentication, leading to the adoption of face recognition (FR) via the cameras embedded in such devices. Progress of deep Convolutional Neural Networks (CNNs) has provided substantial advances in FR. Nonetheless, the size of state-of-the-art architectures is unsuitable for mobile deployment, since they often encompass hundreds of megabytes and millions of parameters. We address this by studying methods for deep network compression applied to FR. In particular, we apply network pruning based on Taylor scores, where less important filters are removed iteratively. The method is tested on three networks based on the small SqueezeNet (1.24M parameters) and the popular MobileNetv2 (3.5M) and ResNet50 (23.5M) architectures. These have been selected to showcase the method on CNNs with different complexities and sizes. We observe that a substantial percentage of filters can be removed with minimal performance loss. Also, filters with the highest amount of output channels tend to be removed first, suggesting that high-dimensional spaces within popular CNNs are over-dimensioned.