ai/ml_十大ML / AI现实世界项目，以增强您的产品组合-CSDN博客

ai/ml

Artificial Intelligence jobs are the sexiest jobs of the century. It is a very frequent topic for movies, books, researchers, and all other media. Every company is planning to invest a lot in it.

人工智能工作是本世纪最性感的工作。对于电影，书籍，研究人员和所有其他媒体来说，这是一个非常常见的话题。每个公司都计划对其进行大量投资。

Did you ever wish you could make an AI project of your own? Well, this article is the best way to get started.

您是否曾经希望过自己做一个AI项目？好吧，本文是入门的最佳方法。

This article will cover many of the real-world ML/AI projects, their problem definition, dataset, and implementation source.

本文将介绍许多现实世界中的ML / AI项目，它们的问题定义，数据集和实现源。

1.人脸识别系统： (1. Face Recognition System:)

Facial Recognition systems are capable of not only recognizing the face of a person but also it can predict gender, age, mood, health condition of the person. Face recognition systems are used across many companies in different sectors such as payments, advertisements, security, social media, etc depending on the company requirements.

面部识别系统不仅能够识别人的面部，而且能够预测人的性别，年龄，情绪，健康状况。人脸识别系统已根据公司的要求在不同行业的许多公司中使用，例如支付，广告，安全性，社交媒体等。

实际的实现： (Real-world implementation:)

Facebook and other social media platforms auto-tags people in the images using their facial recognition systems.
Facebook和其他社交媒体平台使用他们的面部识别系统在图像中为人自动添加标签 。
MasterCard has a selfie pay app called MasterCard Identity Check.
万事达有自拍付费应用 称为万事达卡身份检查。
Gadgets manufacturers like Apple, Samsung use facial recognition as a security feature.
苹果，三星等小工具制造商使用面部识别作为安全功能 。
Grocery giant Tesco plans to install OptimEyes screens to deliver targeted ads to customers.
杂货巨头乐购(Tesco)计划安装OptimEyes屏幕，以向客户投放定向广告 。

Image for post — Source1: Facebook image auto-tagging, 来源1 ：Facebook图片自动标记；来源 Source2: age gender prediction, 2 ：年龄性别预测； Source3: face recognition in atm, (Image edited by Author) 来源3 ：atm中的人脸识别；(图片由作者编辑)

人脸识别项目实施： (Face Recognition project Implementations:)

数据集来源： (Dataset Sources:)

2.垃圾邮件检测和仇恨语音识别： (2. Spam Detection and Identification of Hate Speech:)

Spam Detection and hate speech identification systems detect unwanted, unsolicited, and hate speech text. These systems are used across various social media platforms, articles, emails, forwarded messages, etc.

垃圾邮件检测和仇恨语音识别系统检测到不需要的，未经请求的和仇恨语音文本。这些系统可用于各种社交媒体平台，文章，电子邮件，转发的消息等。

实际的实现： (Real-world implementation:)

Spam filtering used by G-Mail.
G-Mail使用的垃圾邮件过滤。
Hate speech detection in social media such as youtube, twitter, Facebook.
社交媒体(例如youtube，twitter，Facebook)中的仇恨语音检测。

项目实施来源： (Projects Implementations sources:)

数据集来源： (Dataset Sources:)

3.标签预测： (3. Tags Prediction:)

In the world of social media, every article, video, image, etc has tags associated with it. A tag is used in social media to cluster and identify a particular type of material.

在社交媒体世界中，每个文章，视频，图像等都有与其相关的标签。标签在社交媒体中用于聚类和识别特定类型的材料。

实际的实现： (Real-world implementation:)

Tags used in social media such as Twitter, Facebook, Instagram, youtube, etc.
在社交媒体中使用的标签，例如Twitter，Facebook，Instagram，youtube等。
Tags used in a question-answer platform such as StackOverflow, Quora.
在问答平台(如StackOverflow，Quora)中使用的标签。

项目实施来源： (Projects Implementations sources:)

数据集来源： (Dataset Sources:)

4.推荐系统： (4. Recommendation system:)

A recommender system is an information filtering system that seeks to predict the preference of a user would give to an item. They are used across various internet companies to recommend their users.

推荐系统是一种信息过滤系统，旨在预测用户会偏爱的商品。它们在各种互联网公司中用于推荐其用户。

实际的实现： (Real-world implementation:)

Recommendation system on the social media platform like TikTok, Facebook, youtube.
TikTok，Facebook，youtube等社交媒体平台上的推荐系统。
Recommendation system in OTT platforms NetFlix, Hotstar.
OTT平台NetFlix，Hotstar中的推荐系统。
Recommendation system in an e-commerce website like amazon, Walmart.
电子商务网站(如亚马逊，沃尔玛)中的推荐系统。

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

5.聊天机器人： (5. Chatbots:)

A chatbot is a software application used to conduct an online chat conversation via text or text-to-speech, instead of providing direct contact with a live human agent.

聊天机器人是一种软件应用程序，用于通过文本或文本到语音进行在线聊天对话，而不是提供与现场人工代理的直接联系。

实际的实现： (Real-world implementation:)

Chatbots answer to the questions, used by HDFC bank, etc.
聊天机器人回答问题，例如HDFC银行等
Book tickets to event/shows, used by BookMyShow, PVR, etc.
预订活动/演出的票，由BookMyShow，PVR等使用
Chatbots are process return and exchange requests, used by FedEx, Delhivery, etc.
聊天机器人是由FedEx，Delhivery等使用的流程返回和交换请求。

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

6.人类行为和活动识别： (6. Human Action and Activity Recognition:)

Human activity recognition is a time series task that involves predicting the movement of a person based on sensor data such as an accelerometer or gyroscope.

人类活动识别是一个时间序列任务，涉及基于诸如加速计或陀螺仪之类的传感器数据预测人的运动。

实际的实现： (Real-world implementation:)

Used for surveillance purposes using drones or CCTV cameras.
用于使用无人机或CCTV摄像机进行监视。
Used across smartwatches and gadgets to track the health status of the person.
可跨智能手表和小工具使用，以跟踪人员的健康状况。
Used across the gaming industry.
在整个游戏行业中使用。

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

7.手写识别： (7. Handwriting Recognition:)

Handwriting recognition, a category of OCR, is the ability of a computer to receive and interpret intelligible handwritten input and convert to text.

手写识别是OCR的一种，它是计算机接收和解释可理解的手写输入并将其转换为文本的能力。

实际的实现： (Real-world implementation:)

OCR for data entry
OCR数据输入
OCR used in gadgets
小工具中使用的OCR
Validation of signatures
签名验证

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

8.价格预测： (8. Price Prediction:)

Profitable and cheap fare estimation is one of the important business tactics used by various e-commerce, ticket booking websites. An AI model helps these companies to estimate a price. Price prediction systems are also used across various stock trading companies to predict future stock prices.

有利可图的票价估算是各种电子商务，机票预订网站使用的重要商业策略之一。人工智能模型可以帮助这些公司估算价格。价格预测系统还用于各种股票交易公司，以预测未来的股票价格。

实际的实现： (Real-world implementation:)

Product price prediction for various e-commerce websites, such as Amazon, Walmart, etc.
各种电子商务网站(例如Amazon，Walmart等)的产品价格预测
Stock Price Prediction, used by Groww.
股票价格预测，由Groww使用。
Ticket fare prediction for flights, hotels, cabs, used by Airbnb, Oyo, uber, etc.
Airbnb，Oyo，uber等使用的航班，酒店，出租车的票价预测

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

9.音乐流派分类： (9. Music Genre Classification:)

Music genre classification systems are used across various music streaming applications. It classifies music for better user satisfaction. This model can be used to automatically classify the music genre by extracting information from the audio samples such as spectrograms, MFCC, etc.

音乐流派分类系统用于各种音乐流应用程序。它对音乐进行分类，以提高用户满意度。通过从音频样本中提取信息(例如声谱图，MFCC等)，此模型可用于自动分类音乐流派。

实际的实现： (Real-world implementation:)

Used by music streaming platforms such as Spotify, Gaana, etc.
由音乐流平台(例如Spotify，Gaana等)使用。

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

10.图像字幕生成器 (10. Image Caption Generator)

Image captioning is part of NLP. Computers are far behind humans in understanding the context by seeing an image. Image Captioning generator can automatically generate captions for the image.

图像字幕是NLP的一部分。通过查看图像，计算机在理解上下文方面远远落后于人类。图像字幕生成器可以自动为图像生成字幕。

实际的实现： (Real-world implementation:)

Google uses image captioning to better its image search result.
Google使用图像标题来改善其图像搜索结果。
Image gallery websites like google photos can understand the uploaded images using Image captioning systems.
Google图片之类的图片网站网站可以使用图片字幕系统了解上传的图片。

项目实施来源： (Projects Implementation sources:)

数据集来源： (Dataset sources:)

结论： (Conclusion:)

In this article, I have provided you with the best real-world implementation of projects on Machine Learning and Deep Learning. Every category of projects has its respective implementation and dataset sources. Doing any of the above-mentioned projects will strengthen your portfolio.

在本文中，我为您提供了有关机器学习和深度学习项目的最佳现实实现。每个类别的项目都有其各自的实现和数据集来源。完成上述任何项目都会增强您的投资组合。