Meta tags:

Headings (most frequently used words):

efficientnet, contents, compound, scaling, variants, see, also, references, external, links,

Text of the page (most frequently used words):
the (24), #efficientnet (19), and (17), google (12), displaystyle (12), wikipedia (9), 2025 (9), scaling (9), edit (9), search (8), with (8), 2024 (8), 2019 (8), phi (8), this (7), neural (7), models (7), computer (6), 2023 (6), 2022 (6), 2017 (6), network (6), compound (6), multiplier (6), page (5), from (5), networks (5), arxiv (5), depth (5), width (5), resolution (5), gamma (5), beta (5), alpha (5), contents (4), text (4), you (4), was (4), september (4), cleanup (4), all (4), vision (4), also (4), other (4), 2021 (4), 2016 (4), alphago (4), model (4), links (4), quoc (4), for (4), convolutional (4), which (4), image (4), cdot (4), article (4), hide (4), move (4), sidebar (4), view (3), may (3), license (3), pages (3), needing (3), learning (3), learn (3), tensorflow (3), tpu (3), see (3), gemini (3), 2018 (3), mobilenet (3), accuracy (3), cite (3), journal (3), help (3), tan (3), mingxing (3), training (3), nas (3), layers (3), size (3), family (3), increasing (3), baseline (3), are (3), geq (3), tools (3), main (3), language (2), toggle (2), table (2), contact (2), about (2), privacy (2), policy (2), terms (2), using (2), commons (2), categories (2), errors (2), short (2), description (2), wikidata (2), articles (2), software (2), artificial (2), machine (2), retrieved (2), programming (2), computing (2), information (2), differentiable (2), general (2), generative (2), transformer (2), 2015 (2), master (2), automl (2), blog (2), external (2), mixup (2), 2020 (2), randaugment (2), fast (2), optimized (2), references (2), published (2), architecture (2), uses (2), been (2), variants (2), mbconv (2), determined (2), original (2), they (2), that (2), such (2), number (2), dimensions (2), approx (2), aligned (2), appearance (2), upload (2), file (2), changes (2), history (2), read (2), log (2), create (2), account (2), donate (2), menu (2), add, topic, mobile, cookie, statement, statistics, developers, code, conduct, legal, safety, contacts, disclaimers, available, under, additional, apply, site, agree, registered, trademark, non, profit, organization, wikimedia, foundation, inc, use, creative, attribution, sharealike, rendered, parsoid, last, edited, utc, hidden, cs1, missing, periodical, matches, https, org, index, php, title, oldid, 1309177911, portals, technology, mindspore, flux, jax, theano, scikit, keras, pytorch, libraries, spinnaker, memristor, vpu, ipu, hardware, inductive, bias, computational, theory, ricci, calculus, pattern, recognition, neuromorphic, automatic, differentiation, statistical, manifold, geometry, category, workspace, labs, pre, trained, future, summit, attention, need, antigravity, robotics, vids, notebooklm, dreambooth, veo, genie, gemma, videopoet, imagen, palm, chinchilla, lamda, xlnet, bert, nano, banana, sparrow, assistant, chatbots, tensor, processing, unit, quantum, intelligence, lab, gato, wavenet, 2014, inception, alphagenome, alphaevolve, alphaproof, alphageometry, funsearch, alphadev, alphatensor, alphastar, alphafold, popular, culture, jie, lee, sedol, fan, hui, competitions, muzero, alphazero, zero, versions, programs, deepmind, brain, improving, efficiency, through, zhang, hongyi, cisse, moustapha, dauphin, yann, lopez, paz, david, 1710, 09412, beyond, empirical, risk, minimization, requires, cubuk, ekin, zoph, barret, shlens, jonathon, 703, 1909, 13719, 702, practical, automated, data, augmentation, reduced, space, 2104, 00298, efficientnetv2, smaller, faster, sheng, pang, ruoming, andrew, cheng, liqun, jouppi, norman, 2102, 05610, searching, families, datacenter, accelerators, august, research, edgetpu, creating, accelerator, 1905, 11946, rethinking, only, look, once, squeezenet, june, improved, further, more, types, introduced, method, progressively, increases, during, regularization, techniques, like, authors, claim, approach, mitigates, drops, often, associated, progressive, resizing, dropout, has, adapted, inference, centralized, clusters, gpu, tpus, edge, stack, shapes, publication, consisted, subsequent, obtained, architecturally, choice, modules, found, inverted, bottleneck, convolution, called, used, worked, well, introduces, instead, one, dimension, time, channels, input, scale, three, simultaneously, specifically, given, scaled, according, following, equations, subject, condition, factor, would, increase, total, flops, running, approximately, times, small, paper, suggested, respectively, grid, hyperparameters, begin, end, coefficient, have, adopted, various, tasks, including, segmentation, object, detection, classification, cnns, researchers, its, key, innovation, uniformly, scales, single, parameter, github, com, tree, official, repository, website, apache, python, written, release, developer, meet, specific, problem, how, when, remove, message, please, can, improve, citation, formatting, quality, standards, require, free, encyclopedia, item, projects, printable, version, download, pdf, print, export, switch, legacy, parser, get, shortened, url, permanent, link, related, what, here, actions, english, talk, top, personal, special, recent, community, portal, contribute, random, current, events, navigation, jump, content,

Text of the page (random words):
efficientnet wikipedia jump to content main menu main menu move to sidebar hide navigation main page contents current events random article about wikipedia contact us contribute help learn to edit community portal recent changes upload file special pages search search appearance donate create account log in personal tools donate create account log in contents move to sidebar hide top 1 compound scaling 2 variants 3 see also 4 references 5 external links toggle the table of contents efficientnet 1 language 中文 edit links article talk english read edit view history tools tools move to sidebar hide actions read edit view history general what links here related changes upload file permanent link page information cite this page get shortened url switch to legacy parser print export download as pdf printable version in other projects wikidata item appearance move to sidebar hide from wikipedia the free encyclopedia this article may require cleanup to meet wikipedia s quality standards the specific problem is citation formatting errors please help improve this article if you can september 2025 learn how and when to remove this message family of computer vision models efficientnet developer google ai release may 2019 written in python license apache license 2 0 website google ai blog repository github com tensorflow tpu tree master models official efficientnet efficientnet is a family of convolutional neural networks cnns for computer vision published by researchers at google ai in 2019 1 its key innovation is compound scaling which uniformly scales all dimensions of depth width and resolution using a single parameter efficientnet models have been adopted in various computer vision tasks including image classification object detection and segmentation compound scaling edit efficientnet introduces compound scaling which instead of scaling one dimension of the network at a time such as depth number of layers width number of channels or resolution input image size uses a compound coefficient ϕ displaystyle phi to scale all three dimensions simultaneously specifically given a baseline network the depth width and resolution are scaled according to the following equations 1 depth multiplier d α ϕ width multiplier w β ϕ resolution multiplier r γ ϕ displaystyle begin aligned text depth multiplier d alpha phi text width multiplier w beta phi text resolution multiplier r gamma phi end aligned subject to α β 2 γ 2 2 displaystyle alpha cdot beta 2 cdot gamma 2 approx 2 and α 1 β 1 γ 1 displaystyle alpha geq 1 beta geq 1 gamma geq 1 the α β 2 γ 2 2 displaystyle alpha cdot beta 2 cdot gamma 2 approx 2 condition is such that increasing ϕ displaystyle phi by a factor of ϕ 0 displaystyle phi _ 0 would increase the total flops of running the network on an image approximately 2 ϕ 0 displaystyle 2 phi _ 0 times the hyperparameters α displaystyle alpha β displaystyle beta and γ displaystyle gamma are determined by a small grid search the original paper suggested 1 2 1 1 and 1 15 respectively architecturally they optimized the choice of modules by neural architecture search nas and found that the inverted bottleneck convolution which they called mbconv used in mobilenet worked well the efficientnet family is a stack of mbconv layers with shapes determined by the compound scaling the original publication consisted of 8 models from efficientnet b0 to efficientnet b7 with increasing model size and accuracy efficientnet b0 is the baseline network and subsequent models are obtained by scaling the baseline network by increasing ϕ displaystyle phi variants edit efficientnet has been adapted for fast inference on edge tpus 2 and centralized tpu or gpu clusters by nas 3 efficientnet v2 was published in june 2021 the architecture was improved by further nas search with more types of convolutional layers 4 it also introduced a training method which progressively increases image size during training and uses regularization techniques like dropout randaugment 5 and mixup 6 the authors claim this approach mitigates accuracy drops often associated with progressive resizing see also edit convolutional neural network squeezenet mobilenet you only look once references edit 1 2 tan mingxing le quoc v 2020 09 11 efficientnet rethinking model scaling for convolutional neural networks arxiv 1905 11946 efficientnet edgetpu creating accelerator optimized neural networks with automl research google august 6 2019 retrieved 2024 10 18 li sheng tan mingxing pang ruoming li andrew cheng liqun le quoc jouppi norman p 2021 02 10 searching for fast model families on datacenter accelerators arxiv 2102 05610 tan mingxing le quoc v 2021 06 23 efficientnetv2 smaller models and faster training arxiv 2104 00298 cubuk ekin d zoph barret shlens jonathon le quoc v 2020 randaugment practical automated data augmentation with a reduced search space 702 703 arxiv 1909 13719 cite journal cite journal requires journal help zhang hongyi cisse moustapha dauphin yann n lopez paz david 2018 04 27 mixup beyond empirical risk minimization arxiv 1710 09412 external links edit efficientnet improving accuracy and efficiency through automl and model scaling google ai blog v t e google ai google google brain google deepmind computer programs alphago versions alphago 2015 master 2016 alphago zero 2017 alphazero 2017 muzero 2019 competitions fan hui 2015 lee sedol 2016 ke jie 2017 in popular culture alphago 2017 other alphafold 2018 alphastar 2019 alphatensor 2022 alphadev 2023 funsearch 2023 alphageometry 2024 alphaproof 2024 alphaevolve 2025 alphagenome 2025 machine learning neural networks inception 2014 wavenet 2016 mobilenet 2017 transformer 2017 efficientnet 2019 gato 2022 other quantum artificial intelligence lab tensorflow tensor processing unit generative ai chatbots assistant 2016 sparrow 2022 gemini 2023 nano banana 2025 models bert 2018 xlnet 2019 t5 2019 lamda 2021 chinchilla 2022 palm 2022 imagen 2023 gemini 2023 videopoet 2024 gemma 2024 genie 2024 veo 2024 other dreambooth 2022 notebooklm 2023 vids 2024 gemini robotics 2025 antigravity 2025 see also attention is all you need future of go summit generative pre trained transformer google labs google workspace category commons v t e differentiable computing general differentiable programming information geometry statistical manifold automatic differentiation neuromorphic computing pattern recognition ricci calculus computational learning theory inductive bias hardware ipu tpu vpu memristor spinnaker software libraries tensorflow pytorch keras scikit learn theano jax flux jl mindspore portals computer programming technology retrieved from https en wikipedia org w index php title efficientnet oldid 1309177911 categories machine learning computer vision artificial neural networks google software hidden categories articles needing cleanup from september 2025 all pages needing cleanup wikipedia pages needing cleanup from september 2025 articles with short description short description matches wikidata cs1 errors missing periodical this page was last edited on 2 september 2025 at 15 06 utc page was rendered with parsoid text is available under the creative commons attribution sharealike 4 0 license additional terms may apply by using this site you agree to the terms of use and privacy policy wikipedia is a registered trademark of the wikimedia foundation inc a non profit organization privacy policy about wikipedia disclaimers contact wikipedia legal safety contacts code of conduct developers statistics cookie statement mobile view search search toggle the table of contents efficientnet 1 language add topic