# A Powerful Generative Model Using Random Weights for the Deep Image Representation

arXiv: https://arxiv.org/abs/1606.04801 · HTML: https://mlsanity.com/paper/1606.04801/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Jun 2016.

## Built on by (most cited first, 14 of 14)

- [Rethinking and Improving the Robustness of Image Style Transfer](https://mlsanity.com/paper/2104.05623/) (2021)
- [Contrastive Audio-Language Learning for Music](https://mlsanity.com/paper/2208.12208/) (2022)
- [MC-LSTM: Mass-Conserving LSTM](https://mlsanity.com/paper/2101.05186/) (2021)
- [Adversarial Reprogramming Revisited](https://mlsanity.com/paper/2206.03466/) (2022)
- [Generic Perceptual Loss for Modeling Structured Output Dependencies](https://mlsanity.com/paper/2103.10571/) (2021)
- [Neural Architecture Search of Deep Priors: Towards Continual Learning without Catastrophic Interference](https://mlsanity.com/paper/2104.06788/) (2021)
- [I can listen but cannot read: An evaluation of two-tower multimodal systems for instrument recognition](https://mlsanity.com/paper/2407.18058/) (2024)
- [Learning to Act through Evolution of Neural Diversity in Random Neural Networks](https://mlsanity.com/paper/2305.15945/) (2023)
- [Causality-guided Prompt Learning for Vision-language Models via Visual Granulation](https://mlsanity.com/paper/2509.03803/) (2025)
- [DiCOVA-Net: Diagnosing COVID-19 using Acoustics based on Deep Residual Network for the DiCOVA Challenge 2021](https://mlsanity.com/paper/2107.06126/) (2021)
- [Gradient Kernel Regression](https://mlsanity.com/paper/2104.05874/) (2021)
- [Random Embeddings and Linear Regression can Predict Protein Function](https://mlsanity.com/paper/2104.14661/) (2021)
- [Untrained Perceptual Loss for image denoising of line-like structures in MR images](https://mlsanity.com/paper/2411.05884/) (2024)
- [Can Impressions of Music be Extracted from Thumbnail Images?](https://mlsanity.com/paper/2501.02511/) (2025)

## Abstract (arXiv)

To what extent is the success of deep visualization due to the training? Could we do deep visualization using untrained, random weight networks? To address this issue, we explore new and powerful generative models for three popular deep visualization tasks using untrained, random weight convolutional neural networks. First we invert representations in feature spaces and reconstruct images from white noise inputs. The reconstruction quality is statistically higher than that of the same method applied on well trained networks with the same architecture. Next we synthesize textures using scaled correlations of representations in multiple layers and our results are almost indistinguishable with the original natural texture and the synthesized textures based on the trained network. Third, by recasting the content of an image in the style of various artworks, we create artistic images with high perceptual quality, highly competitive to the prior work of Gatys et al. on pretrained networks. To our knowledge this is the first demonstration of image representations using untrained deep neural networks. Our work provides a new and fascinating tool to study the representation of deep network architecture and sheds light on new understandings on deep visualization.