[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-3877":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":16,"forks30d":16,"starsTrendScore":16,"compositeScore":19,"rankGlobal":10,"rankLanguage":10,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":21,"hasPages":21,"topics":23,"createdAt":10,"pushedAt":10,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":16,"starSnapshotCount":16,"syncStatus":27,"lastSyncTime":28,"discoverSource":29},3877,"tfjs-models","tensorflow\u002Ftfjs-models","tensorflow","Pretrained models for TensorFlow.js","https:\u002F\u002Fjs.tensorflow.org",null,"TypeScript",14796,4382,275,261,0,3,18,45,"Apache License 2.0",false,"master",[],"2026-06-12 02:00:55","# Pre-trained TensorFlow.js models\n\nThis repository hosts a set of pre-trained models that have been ported to\nTensorFlow.js.\n\nThe models are hosted on NPM and unpkg so they can be used in any project out of the box. They can be used directly or used in a transfer learning\nsetting with TensorFlow.js.\n\nTo find out about APIs for models, look at the README in each of the respective\ndirectories. In general, we try to hide tensors so the API can be used by\nnon-machine learning experts.\n\nFor those interested in contributing a model, please file a [GitHub issue on tfjs](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Ftfjs\u002Fissues) to gauge\ninterest. We are trying to add models that complement the existing set of models\nand can be used as building blocks in other apps.\n\n## Models\n\n\u003Ctable style=\"max-width:100%;table-layout:auto;\">\n  \u003Ctr style=\"text-align:center;\">\n    \u003Cth>Type\u003C\u002Fth>\n    \u003Cth>Model\u003C\u002Fth>\n    \u003Cth>Demo\u003C\u002Fth>\n    \u003Cth>Details\u003C\u002Fth>\n    \u003Cth>Install\u003C\u002Fth>\n  \u003C\u002Ftr>\n  \u003C!-- Images -->\n  \u003C!-- ** MobileNet -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"12\">\u003Cb>Images\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fmobilenet\">\u003Cdiv style='vertical-align:middle; display:inline;'>MobileNet\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002Fmobilenet\u002Findex.html\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Classify images with labels from the \u003Ca href=\"http:\u002F\u002Fwww.image-net.org\u002F\">ImageNet database\u003C\u002Fa>.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fmobilenet\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fmobilenet\u002Fdemo\u002Findex.html\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C!-- ** Hand -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fhand-pose-detection\">\u003Cdiv style='vertical-align:middle; display:inline;'>Hand\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002Fhand-pose-detection\u002Findex.html?model=mediapipe_hands\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Real-time hand pose detection in the browser using TensorFlow.js.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fhand-pose-detection\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fhand-pose-detection\u002Fdemos\u002Flive_video\u002Findex.html\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n    \u003C!-- ** Pose -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fpose-detection\">\u003Cdiv style='vertical-align:middle; display:inline;'>Pose\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002Fpose-detection\u002Findex.html?model=movenet\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">An API for real-time human pose detection in the browser.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fpose-detection\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fpose-detection\u002Fdemos\u002Flive_video\u002Findex.html\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C!-- ** Coco SSD -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fcoco-ssd\">\u003Cdiv style='vertical-align:middle; display:inline;'>Coco SSD\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"\">\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Object detection model that aims to localize and identify multiple objects in a single image. Based on the \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels\u002Fblob\u002Fmaster\u002Fresearch\u002Fobject_detection\u002FREADME.md\">TensorFlow object detection API\u003C\u002Fa>.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fcoco-ssd\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fcoco-ssd\u002Fdemo\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fdeeplab\">\u003Cdiv style='vertical-align:middle; display:inline;'>DeepLab v3\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"\">\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Semantic segmentation\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fdeeplab\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fdeeplab\u002Fdemo\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n    \u003C!-- ** Face Landmark Detection -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fface-landmarks-detection\">\u003Cdiv style='vertical-align:middle; display:inline;'>Face Landmark Detection\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002Fface-landmarks-detection\u002Findex.html?model=mediapipe_face_mesh\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Real-time 3D facial landmarks detection to infer the approximate surface geometry of a human face\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fface-landmarks-detection\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fface-landmarks-detection\u002Fdemos\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n\n  \u003C!-- * Audio -->\n  \u003C!-- ** Speech Commands -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>Audio\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fspeech-commands\">\u003Cdiv style='vertical-align:middle; display:inline;'>Speech Commands\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-speech-model-test\u002F2019-01-03a\u002Fdist\u002Findex.html\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Classify 1 second audio snippets from the \u003Ca href=\"https:\u002F\u002Fwww.tensorflow.org\u002Ftutorials\u002Faudio\u002Fsimple_audio\">speech commands dataset\u003C\u002Fa>.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fspeech-commands\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fspeech-commands\u002Fdemo\u002Findex.html\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C!-- * Text -->\n  \u003C!-- ** Universal Sentence Encoder -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"4\">\u003Cb>Text\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Funiversal-sentence-encoder\">\u003Cdiv style='vertical-align:middle; display:inline;'>Universal Sentence Encoder\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"\">\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Encode text into a 512-dimensional embedding to be used as inputs to natural language processing tasks such as sentiment classification and textual similarity.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Funiversal-sentence-encoder\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Funiversal-sentence-encoder\u002Fdemo\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C!-- ** Text Toxicity -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Ftoxicity\">\u003Cdiv style='vertical-align:middle; display:inline;'>Text Toxicity\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002Ftoxicity\u002Findex.html\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Score the perceived impact a comment might have on a conversation, from \"Very toxic\" to \"Very healthy\".\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Ftoxicity\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Ftoxicity\u002Fdemo\u002Findex.html\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C!-- * Depth Estimation -->\n  \u003C!-- ** Portrait Depth -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>Depth Estimation\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fdepth-estimation\">\u003Cdiv style='vertical-align:middle; display:inline;'>Portrait Depth\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002F3dphoto\u002Findex.html\">live\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">Estimate per-pixel depth (the distance to the camera center) for a single portrait image, which can be further used for creative applications such as \u003Ca href=\"https:\u002F\u002Fblog.tensorflow.org\u002F2022\u002F05\u002Fportrait-depth-api-turning-single-image.html?linkId=8063793\">3D photo\u003C\u002Fa> and \u003Ca href=\"https:\u002F\u002Fstorage.googleapis.com\u002Ftfjs-models\u002Fdemos\u002Frelighting\u002Findex.html\">relighting\u003C\u002Fa>.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fdepth-estimation\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fdepth-estimation\u002Fdemos\u002F3d_photo\u002Findex.html\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C!-- * General Utilities -->\n  \u003Ctr>\n    \u003Ctd rowspan=\"2\">\u003Cb>General Utilities\u003C\u002Fb>\u003C\u002Ftd>\n  \u003C!-- ** KNN Classifier -->\n    \u003Ctd rowspan=\"2\">\u003Cb>\u003Ca style=\"white-space:nowrap; display:inline-block;\" href=\".\u002Fknn-classifier\">\u003Cdiv style='vertical-align:middle; display:inline;'>KNN Classifier\u003C\u002Fdiv>\u003C\u002Fa>\u003C\u002Fb>\u003C\u002Ftd>\n    \u003Ctd>\u003Ca href=\"\">\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">This package provides a utility for creating a classifier using the K-Nearest Neighbors algorithm. Can be used for transfer learning.\u003C\u002Ftd>\n    \u003Ctd rowspan=\"2\">\u003Ccode>npm i @tensorflow-models\u002Fknn-classifier\u003C\u002Fcode>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Ca href=\".\u002Fknn-classifier\u002Fdemo\">source\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n\u003C\u002Ftable>\n\n## Development\n\nYou can run the unit tests for any of the models by running the following\ninside a directory:\n\n`yarn test`\n\nNew models should have a test NPM script (see [this](.\u002Fmobilenet\u002Fpackage.json) `package.json` and `run_tests.ts` [helper](.\u002Fmobilenet\u002Frun_tests.ts) for reference).\n\nTo run all of the tests, you can run the following command from the root of this\nrepo:\n\n`yarn presubmit`\n","tensorflow\u002Ftfjs-models 是一个为 TensorFlow.js 提供预训练模型的项目。该项目包含多种预训练模型，如图像分类、手部姿态检测和人体姿态检测等，这些模型可以直接通过 NPM 或 unpkg 获取并在任何项目中使用，支持直接应用或在迁移学习场景下使用。为了方便非机器学习专家使用，项目的 API 设计尽量简化了张量操作。适用于需要快速集成高质量机器学习模型的 Web 应用开发场景，比如在线图像识别、实时姿态追踪等功能实现。",2,"2026-06-11 02:56:51","top_language"]