ResNet
In plain wordsA family of image-recognition networks from 2015 whose shortcut connections allowed far deeper networks than before, over 100 layers, and became a standard building block across AI.
A convolutional neural networkconvolutional neural network, n.: A kind of neural network built for images and other grid-like data. It slides small pattern detectors across the input, finding edges first, then shapes, then whole objects. architecture built from residual blocks, each computing x + F(x), where F is two 3 × 3 convolutionsconvolution, n.: An operation that slides a small grid of weights across an image or signal, multiplying and adding at each spot, to measure how strongly a pattern appears at every location. or, in deeper versions, a 1 × 1–3 × 3–1 × 1 'bottleneck', with batch normalizationbatch normalization, n.: A layer that rescales the numbers flowing through a network so that, across each batch of training examples, they keep a steady average and spread. It makes deep networks train faster and more reliably. and ReLUReLU, abbr.: The most common rule a neuron uses to produce its output: keep the number if it is positive, replace it with zero if it is negative. Simple, fast and surprisingly effective. (residual connectionresidual connection, n.: A shortcut in a neural network that adds a block's input straight to its output, so each block only has to learn a correction. It is what makes networks hundreds of layers deep trainable.). Standard depths are 18, 34, 50, 101 and 152 layers; ResNet-50, with about 25 million parameters, became the default backbone for classification, detection and segmentation and a common benchmark for hardware and transfer learningtransfer learning, n.: Reusing what a model learned on one task to get a head start on another — like a pianist picking up the organ faster than a beginner. It lets models learn new jobs from little data.. Downsampling uses stridedstride, n.: How far a scanning window jumps each time it moves across an image. Jumping two pixels instead of one halves the size of the result. convolutions with projection shortcuts, and a global-average-poolingpooling, n.: A step in an image network that shrinks its internal picture by summarizing each small patch with one number, usually the largest or the average, keeping what matters and discarding exact positions. head replaces large fully connected layers. Variants include pre-activation ResNets (2016), Wide ResNets and ResNeXt; ConvNeXt (2022) modernized the design to compete with vision transformersvision transformer, n.: A way of applying the transformer, the design behind chatbots, to pictures: the image is cut into small square tiles that the model reads like words in a sentence..
We use an ImageNet-pretrained ResNet-50 as the backbone and fine-tune all layers.
Origin
Introduced by Kaiming He, Xiangyu Zhang, Shaoqing Ren and Jian Sun of Microsoft Research in 'Deep Residual Learning for Image Recognition' (arXiv, December 2015; CVPR 2016 best paper). An ensemble won the ILSVRC 2015 classification task with 3.57% top-5 test error, and ResNets also won the ILSVRC 2015 detection and localization tasks and the COCO 2015 detection and segmentation tasks. An April 2025 Nature analysis ranked the paper the most-cited of the 21st century.
Related
- residual connectionresidual connection, n.: A shortcut in a neural network that adds a block's input straight to its output, so each block only has to learn a correction. It is what makes networks hundreds of layers deep trainable.
- convolutional neural networkconvolutional neural network, n.: A kind of neural network built for images and other grid-like data. It slides small pattern detectors across the input, finding edges first, then shapes, then whole objects.
- batch normalizationbatch normalization, n.: A layer that rescales the numbers flowing through a network so that, across each batch of training examples, they keep a steady average and spread. It makes deep networks train faster and more reliably.
- highway networkhighway network, n.: An early design for very deep networks in which learned gates let information either pass through a layer unchanged or be transformed, making networks hundreds of layers deep trainable.
- ImageNetImageNet, prop. n.: A huge collection of more than 14 million labeled photos, sorted into object categories, whose yearly recognition contest from 2010 to 2017 sparked the deep learning boom.
- vision transformervision transformer, n.: A way of applying the transformer, the design behind chatbots, to pictures: the image is cut into small square tiles that the model reads like words in a sentence.
Sources
- He, K., Zhang, X., Ren, S. & Sun, J. (2015). Deep Residual Learning for Image Recognition. arXiv:1512.03385; CVPR 2016. arxiv.org/abs/1512.03385
- Nature (2025). Exclusive: the most-cited papers of the twenty-first century. Nature news feature, April 2025. nature.com/articles/d41586-025-01125-9
Mentioned in
image classificationimage classification, n.: Teaching a computer to say what a picture shows by choosing a label, such as 'cat', 'pneumonia' or 'stop sign', from a fixed list of possibilities., AlphaGo ZeroAlphaGo Zero, prop. n.: A program from DeepMind that taught itself to play the board game Go better than any human in 2017, starting from nothing but the rules and learning only by playing millions of games against itself., architecturearchitecture, n.: The overall design of an AI model, meaning what kinds of parts it has and how they are connected, as opposed to the specific numbers it learns during training., backbonebackbone, n.: The main, usually pretrained part of a vision model that turns an image into useful internal features; smaller task-specific parts are attached to it, like limbs to a spine., batch normalizationbatch normalization, n.: A layer that rescales the numbers flowing through a network so that, across each batch of training examples, they keep a steady average and spread. It makes deep networks train faster and more reliably., CLIPCLIP, prop. n.: An OpenAI model that learned to match pictures with their captions from hundreds of millions of web examples, so it can judge how well any piece of text describes any image., convolutional neural networkconvolutional neural network, n.: A kind of neural network built for images and other grid-like data. It slides small pattern detectors across the input, finding edges first, then shapes, then whole objects., CVPRCVPR, prop. n.: The leading yearly conference on computer vision — teaching computers to understand images and video — and one of the most cited publication venues in any field of science., Kaiming HeKaiming He, prop. n.: A Chinese computer scientist whose ResNet design, with shortcut connections that let networks grow hundreds of layers deep, is among the most cited research in all of science., Microsoft ResearchMicrosoft Research, prop. n.: Microsoft's research division, with laboratories on several continents, whose scientists have produced influential work on speech, vision and language AI., poolingpooling, n.: A step in an image network that shrinks its internal picture by summarizing each small patch with one number, usually the largest or the average, keeping what matters and discarding exact positions., residual connectionresidual connection, n.: A shortcut in a neural network that adds a block's input straight to its output, so each block only has to learn a correction. It is what makes networks hundreds of layers deep trainable., stridestride, n.: How far a scanning window jumps each time it moves across an image. Jumping two pixels instead of one halves the size of the result., ConvNeXtConvNeXt, prop. n.: A 2022 image-recognition network showing that an older-style convolutional design, updated with lessons from transformers, can match vision transformers., DenseNetDenseNet, prop. n.: An image-recognition network in which every layer receives the outputs of all the layers before it, so features are reused rather than relearned., highway networkhighway network, n.: An early design for very deep networks in which learned gates let information either pass through a layer unchanged or be transformed, making networks hundreds of layers deep trainable., InceptionInception, prop. n.: A family of Google image-recognition networks, the first called GoogLeNet, whose building blocks examine each patch of an image at several scales at once while keeping computation low., squeeze-and-excitation networksqueeze-and-excitation network, n.: An image-network design whose add-on blocks let the network decide, for each picture, which of its internal feature channels to turn up and which to turn down., VGGNetVGGNet, prop. n.: An influential 2014 image-recognition network from Oxford, built by stacking many identical small pattern detectors, which showed that making networks deeper by itself improves accuracy.