mlsanityData to 14 Oct 2026

Cut Inner Layers: A Structured Pruning Strategy for Efficient U-Net GANs

arXiv 2206.14658 ↗First posted 29 Jun 2022cs.LG cs.CVcited by 5

Not on the map. Why?

Abstract (arXiv)

Pruning effectively compresses overparameterized models. Despite the success of pruning methods for discriminative models, applying them for generative models has been relatively rarely approached. This study conducts structured pruning on U-Net generators of conditional GANs. A per-layer sensitivity analysis confirms that many unnecessary filters exist in the innermost layers near the bottleneck and can be substantially pruned. Based on this observation, we prune these filters from multiple inner layers or suggest alternative architectures by completely eliminating the layers. We evaluate our approach with Pix2Pix for image-to-image translation and Wav2Lip for speech-driven talking face generation. Our method outperforms global pruning baselines, demonstrating the importance of properly considering where to prune for U-Net generators.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 35 of its 54 references that are in the map. First posted on arXiv 29 Jun 2022.

Builds on

35 of 54 references.

References · 35

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  3. 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,527
  4. 2018The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networkscited by 2,262
  5. 2014FitNets: Hints for Thin Deep Netscited by 1,653
  6. 2016Pruning Filters for Efficient ConvNetscited by 1,143
  7. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  8. 2020A Lip Sync Expert Is All You Need for Speech to Lip Generation In The Wildcited by 672
  9. 2016Learning Structured Sparsity in Deep Neural Networkscited by 541
  10. 2019The State of Sparsity in Deep Neural Networkscited by 476
  11. 2017Efficient Processing of Deep Neural Networks: A Tutorial and Surveycited by 434
  12. 2020Reliable Fidelity and Diversity Metrics for Generative Modelscited by 434
23 more
  1. 2018Filter Pruning via Geometric Median for Deep Convolutional Neural Networks Accelerationcited by 367
  2. 2018LRS3-TED: a large-scale dataset for visual speech recognitioncited by 308
  3. 2020Towards Photo-Realistic Virtual Try-On by Adaptively GeneratingleftrightarrowPreserving Image Contentcited by 149
  4. 2020Layer-adaptive sparsity for the Magnitude-based Pruningcited by 126
  5. 2019AutoGAN: Neural Architecture Search for Generative Adversarial Networkscited by 94
  6. 2019Pruning by Explaining: A Novel Criterion for Deep Neural Network Pruningcited by 89
  7. 2020AutoGAN-Distiller: Searching to Compress Generative Adversarial Networkscited by 50
  8. 2020DSA: More Efficient Budgeted Pruning via Differentiable Sparsity Allocationcited by 47
  9. 2020GAN Slimming: All-in-One GAN Compression by A Unified Optimization Frameworkcited by 44
  10. 2019AdversarialNAS: Adversarial Neural Architecture Search for GANscited by 38
  11. 2021Teachers Do More Than Teach: Compressing Image-to-Image Modelscited by 37
  12. 2019Co-Evolutionary Compression for Unpaired Image Translationcited by 32
  13. 2021GANs Can Play Lottery Tickets Toocited by 32
  14. 2021Content-Aware GAN Compressioncited by 28
  15. 2020Slimmable Generative Adversarial Networkscited by 27
  16. 2021Data-Efficient GAN Training Beyond (Just) Augmentations: A Lottery Ticket Perspectivecited by 25
  17. 2021Revisiting Discriminator in GAN Compression: A Generator-discriminator Cooperative Compression Schemecited by 23
  18. 2019QGAN: Quantized Generative Adversarial Networkscited by 20
  19. 2020Learning Efficient GANs for Image Translation via Differentiable Masks and co-Attention Distillationcited by 17
  20. 2019Localization-aware Channel Pruning for Object Detectioncited by 13
  21. 2021Automatic Correction of Internal Units in Generative Neural Networkscited by 8
  22. 2021Quantization of Generative Adversarial Networks for Efficient Inference: a Methodological Studycited by 4
  23. 2021Evaluating Post-Training Compression in GANs using Locality-Sensitive Hashingcited by 1

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 3; 2024: 0; 2025: 1.

5 papers.

  1. 2023BK-SDM: A Lightweight, Fast, and Cheap Version of Stable Diffusioncited by 13
  2. 2022Discriminator-Cooperated Feature Map Distillation for GAN Compressioncited by 4
  3. 2023Compressing Image-to-Image Translation GANs Using Local Density Structures on Their Learned Manifoldcited by 3
  4. 2023A Unified Compression Framework for Efficient Speech-Driven Talking-Face Generationcited by 0
  5. 2025KDC-Diff: A Latent-Aware Diffusion Model with Knowledge Retention for Memory-Efficient Image Generationcited by 0