2c:null 2d:["$","title",null,{"children":"SM3809_A10 Machine Learning I | stevenzhou01 | Observable"}] 32:T4e0,- Usage: ``` { const bodypix = ml5.bodyPix(modelReady); function modelReady() { // segment the image given bodypix.segment(img, gotResults); } function gotResults(error, result) { if (error) { console.log(error); return; } // log the result console.log(result.backgroundMask); } } ``` (1) Initialize:
const bodypix = ml5.bodyPix(video, options, callback);
Video: an HTMLVideoElement (optional);
Options: an object to change the default values (optional);
Callback: a function runs when the model is loaded. For here, modelReady is a property return boolean value to indicate if the model has been loaded.

(2) Methods:
bodypix.segment(input, options, callback); allows you to segment a person from the background.
Input: image or video data, the input here is "img"
Options: an object to change the default values (optional);
Callback: A function to handle the results of .segment(), here is "gotResults".
Output: either an array of 8-bit unsigned integers or p5Image.
bodyPix.segmentWithParts(input, options, callback); is similar to .segment(), which allows you to get the segmented body parts of the person 33:[{"id":"c1ec97ff3b9252349eaa6e2b452b59ff9ceaa3440ce80b2f7b0342cb0d1621f1d1865987a1da6ffbed9abddd8b24de7bb9d8f833cb07b95a5b0af1657a3e02cc","href":"https://static.observableusercontent.com/files/c1ec97ff3b9252349eaa6e2b452b59ff9ceaa3440ce80b2f7b0342cb0d1621f1d1865987a1da6ffbed9abddd8b24de7bb9d8f833cb07b95a5b0af1657a3e02cc","name":"bodypix2-twitter.gif","type":"image/gif","lastModified":1648714182475.322,"size":10659492}] 30:{"id":"9785cad4959d20d6","slug":"sm3809_a10-machine-learning-i","title":"SM3809_A10 Machine Learning I","thumbnail":"18f6d3e9bdc2c7bb77cc9d317bd64f863a557f5c78ed2b3f5bc8ea7b6d5110da","access_level":"public","create_time":"$D2022-03-31T05:58:24.582Z","update_time":"$D2022-03-31T08:22:08.277Z","view_time":null,"first_public_time":"$D2022-03-31T08:22:27.626Z","has_unpublished_edits":null,"fork":false,"stars":0,"starred":false,"owner":{"type":"team","id":"97db974a0b88bc30","login":"stevenzhou01","name":"","avatar_url":"https://avatars-next.observableusercontent.com/1e2dd4f7120a4e62cca461b2d703afd96e08ebd45e344b3a342ca62b0cfd87e2","tier":"starter_2024"},"version":72,"publish_version":72,"role":null,"authored":false,"authors":[{"type":"individual","id":"f37ed0ff88be839d","login":"stevenzhou01","name":"","avatar_url":"https://avatars-next.observableusercontent.com/1e2dd4f7120a4e62cca461b2d703afd96e08ebd45e344b3a342ca62b0cfd87e2","tier":null}],"members":[{"type":"individual","id":"f37ed0ff88be839d","login":"stevenzhou01","name":"","avatar_url":"https://avatars-next.observableusercontent.com/1e2dd4f7120a4e62cca461b2d703afd96e08ebd45e344b3a342ca62b0cfd87e2","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":1,"collection":{"id":"59eef647f28a2708","slug":"software-art","title":"Software art","owner":{"id":"97db974a0b88bc30","login":"stevenzhou01","name":""}},"type":"notebook","body":{"titling":"auto","title":"SM3809_A10 Machine Learning I","cells":[{"id":0,"value":"# SM3809_A10 Machine Learning I","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":8,"value":"## 1. Reading","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":13,"value":"- A Beginner’s Guide to Neural Networks: Part One
\nThe neural network simulates how human brains work. The Perceptron is like digitalizing our perception and cognition. Human perceives information by senses, then process them through cognition. The training process is like forming the cognition of a machine
\n- Kyle Booten's Sentiment: A Lab Notebook
\nIt's a very informative article. I came up with another application of sentiment analysis. This analysis is widely used in public relations or reputation management in the industry. Companies use this to analyze the press release and the brand image in the public.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":20,"value":"## 2. [BodyPix](https://learn.ml5js.org/#/reference/bodypix?id=description)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":24,"value":"- Intro:
\nDeveloper: \"Bodypix is an open-source machine learning model which allows for person and body-part segmentation in the browser with TensorFlow.js.\" \"TensorFlow.js\" is a library for machine learning in JavaScript, which develops machine learning models in JavaScript, and use it directly in the browser or in Node. Body-part segmentation is a kind of image segmentation, which is used to group pixels of an image in order to locate boundaries and objects. BodyPix is trained to do this for 24 body parts.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":30,"value":"$32","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":42,"value":"- Model Overview
\nMobileNet is a type of machine learning model architecture that could run on platforms with limited computational power, such as mobiles. MobileNet includes the usages of image classification, object detection, and image segmentation. This is also a type of convolutional neural network (CNN). ml5 defaults using a MobileNet created with TensorFlow.js","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":61,"value":"This technique could be used to talk about appearance-related topics, such as gender and races. Use the body-part segmentation to gain the value of different people's body parts, then use the values to regenerate a virtual human. The generative human body could be used to discuss humans are all same regardless of races, genders, or appearance.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":49,"value":"BodyPix 2.0 is also released: Multi-person support and improved accuracy (based on ResNet50) in BodyPix 2.0
\nPress release: [BodyPix: Real-time Person Segmentation in the Browser with TensorFlow.js](https://blog.tensorflow.org/2019/11/updated-bodypix-2.html)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":54,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":65,"value":"## 3. Final Project","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":69,"value":"Documentation link: https://observablehq.com/@stevenzhou01/untitled","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"}],"files":"$W33","stdlib":"1"}} 31:"$30:body"