[HMS Core] HMS Core | How ML Kit Enriches Your Apps with Everyday Intelligence

Smart apps have improved our lives in countless ways, streamlining complex interactions with elegant solutions, such as face-scan payment, product visual search, nuanced real-time translations, and a myriad of other features. All of these life-altering services are made possible by apps that have integrated AI capabilities. HUAWEI ML Kit brings everyday intelligence to a whole new level, offering a treasure trove of next-level visual and language service APIs, and revolutionizing daily life in the process.

01 About HUAWEI ML Kit

HUAWEI ML Kit allows for your apps to leverage Huawei's accumulated prowess in machine learning to support diverse applications for AI. Some prominent examples of Huawei's solutions put into practice include face detection, text recognition, image classification, object detection and tracking, landmark recognition, image segmentation, product visual search, and translation.

02  Typical Services

1. Face Detection

If you are developing an app that incorporates identity authentication, beautification, childcare services, or related recommendations, ML Kit can be a godsend. It is able to detect face orientation, determine whether the user's eyes are open or closed, and identify seven common facial expressions, in addition to other information such as age, gender, and wearing status. ML Kit is also capable of scanning multiple faces simultaneously, and mapping the contours and attributes (nose, eyes, lips, eyebrows, etc.) of each detected face on an 845-point grid, within 50 milliseconds. Better yet, the face detection function is entirely free of charge, and supported on all Android phones.

2. Text Recognition

You might encounter text within an image that you would like to translate, or input into a document, but have no other option than to type it letter-by-letter. ML Kit provides a text recognition API that is capable of extracting text from such items as bank cards, membership cards, or documents, and converting the text to the user's phone, where it can be edited. ML Kit provides for accurate identification, even when the text is tilted at a 30-degree, or even 45-degree angle. This stunning degree of accuracy applies to Simplified Chinese, Japanese, Korean, and all Latin alphabet-based languages, when the text recognition function is running on the device. Even more languages are supported when it is running on the cloud, including Simplified Chinese, English, Spanish, Portuguese, Italian, German, French, Russian, Japanese, Korean, Polish, Finnish, Norwegian, Swedish, Danish, Turkish, Thai, Arabic, and Hindi – remarkable coverage that meets the needs of even the most diverse user base!

3. Image Classification

ML Kit classifies elements in images into intuitive categories, such as individuals, objects, environments, activities, and artwork, to define common themes and application scenarios. This lays the foundation for highly-functional apps that are capable of identifying, classifying, and managing images with remarkable ease and precision. If you have a photo gallery app, integrating it with ML Kit provides for a new layer of intelligence, as images can be automatically sorted by category.

4. Object Detection and Tracking

ML Kit can detect up to eight objects in each image, determine their respective locations within the image, and analyze this data in real time. Remarkably, it can also track objects. Detectable objects fall into the following seven categories: household items, fashion goods, individuals, food, locations, plants, and others. ML Kit substantially outperforms other object detection and tracking services in the market, with regard to both speed and accuracy. Image filter apps based on ML Kit could hardly be easier to develop, or more functional, enabling users to classify everyday images, and filter out objects by category, to zero in on items of interest. It also works in the following manner: Your app passes along a picture taken by the user to ML Kit; ML then detects the object in the image, a handbag for instance, and redirects the user to the online store that sells this type of handbag.

5. Landmark Recognition

ML Kit is capable of recognizing more than 17,000 landmarks across the globe, and obtaining the name and geographic coordinates for each landmark, providing for newfound joy and convenience during trips. If you have developed a travel app, you can make full use of the landmark recognition information to provide users with highly-personalized services. For example, when detecting an image passed by a user, your app will recognize the landmark in it, inform the user where it can be found, and push related information.

6. Image Segmentation

The ML Kit image segmentation function allows for effortless image cropping, empowering users to change the image background with just a click. This feature supports 11 types of scenes/objects, including individuals, skyscrapers, plants, bodies

of water, buildings, sand, mountains, cat and dogs, food, flowers, and others. You can embed 1C portrait mode or 10C mode into your app as needed, and leverage granular

segmentation to an astonishing level of detail at ±2 pixels.

7. Product Visual Search

The customized product visual search service on ML Kit enables app users to purchase an item directly, after taking a picture of it. The search is conducted within an image library that you are free to generate from images of your own products. When the user provides a picture they have taken or scanned to your app, ML Kit will search through the image library for similar items, and when found, return the corresponding product IDs and images. Furthermore, the returned product information can serve as the basis for a whole host of other services, such as similar product comparisons, and searches for imported products or products from celebrity pictures.

8. Translation

There is soaring demand for reliable, real time translations, such as snap to translate and one-click document/web page translation. Translation apps that cater to travelers, or are used for educational purposes, and web browsing, can benefit greatly from integrating ML Kit, which provides for automatic source language detection, and translation between seven languages (Simplified Chinese, English, French, Arabic, Thai, Spanish, and Turkish).

02  Streamlined Integration Process

ML Kit provides all necessary API-based device-cloud synergy services on a single platform, and can run on either a device or on the cloud. On-device ML Kit is scorching-fast, and does not require a network connection, while on-cloud ML Kit provides a remarkable degree of accuracy, thanks to the support of powerful computing, and offers the best solution for complex scenarios. ML Kit supports a wide range of device types, including smartphones, tablets, and smart screens, serving as an indispensable distribution hub, linking your apps with Huawei's immense global user base.

For guidance during the ML Kit integration process, please go to the HUAWEI Developer website, where you will find the integration guide and other resources for reference, or you can contact us by work order for further technical assistance.