# Learning to See by Looking at Noise Source: [Learning to See by Looking at Noise]() Collection index: [llms.txt]() Data updated: 2025-10-26T00:09:48.320Z Authors: Manel Baradad, Jonas Wulff, Tongzhou Wang, Phillip Isola, Antonio Torralba Published: 2021-06-10T17:56:46.000Z ## Abstract Current vision systems are trained on huge datasets, and these datasets come with costs: curation is expensive, they inherit human biases, and there are concerns over privacy and usage rights. To counter these costs, interest has surged in learning from cheaper data sources, such as unlabeled images. In this paper we go a step further and ask if we can do away with real image datasets entirely, instead learning from noise processes. We investigate a suite of image generation models that produce images from simple random processes. These are then used as training data for a visual representation learner with a contrastive loss. We study two types of noise processes, statistical image models and deep generative models under different random initializations. Our findings show that it is important for the noise to capture certain structural properties of real data but that good performance can be achieved even with processes that are far from realistic. We also find that diversity is a key property to learn good representations. Datasets, models, and code are available at https://mbaradad.github.io/learning\_with\_noise. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Retrieval on ImageNet-100 (val)](): Image Retrieval; ImageNet-100 (val); Top-1 Accuracy: 33. ## Original sources and implementation [Original paper]() [PDF]()