Core demands
In the marketing activities of auto enterprises, auto exhibitors and 4S stores, it is difficult for the staff to explain to each user one by one. In order to improve the user's car buying experience and save labor costs, they have put forward the demands of electronic car manuals; On the other hand, car owners and potential buyers also need a tool to take photos and read maps, so that they can more easily obtain information such as vehicle introductions and user manuals.
Solution
The car identification tool developed by Skirt Technology is connected with Baidu image recognition and image search technology to create the first auto electronic manual tool software enabled by AI technology in China. Auto enterprises, auto show organizers and 4S stores can create exclusive auto electronic manuals by uploading a small number of vehicle pictures according to business needs; For car buyers/potential users, in any scene such as street, auto show, 4S store, etc., you can take photos to obtain the price, performance and other information of the vehicle, and quickly understand the details of the vehicle.
1. Vehicle identification
Access Baidu's model identification service, map the model returned from the interface with the self owned model database of Autobab, and associate the price, configuration and other description information of the model.
User usage process:
Step 1: Enter the car identification app to take/upload pictures of car models
Step 2: Identify the model and obtain the model introduction information, such as brand, model, price, basic configuration, etc
Step 3: Follow up operations according to the identification results, such as learning details and booking a test drive
2. Component identification
Based on Baidu's similarity map search technology, Zhichebao provides customized templates for B-end users such as auto companies, auto show organizers, 4S stores, etc. The merchants upload pictures of the whole vehicle and auto parts as required, and add corresponding text, pictures, and video introductions to the models and various parts to generate the electronic manuals specific to the models.
After the creation is successful, the merchant staff will share the page with the car buyers or potential users. The user takes/uploads the photos of the car parts. The back-end calls Baidu similar image search interface, retrieves the pictures of similar parts in the merchant's exclusive image library, returns the introduction information related to the parts, and helps the user understand the details of the car.
The use process is as follows:
Step 1: Merchants shoot and create models (e.g. Audi A4L), and fill in the introduction
Step 2: Create components (such as instrument panel, central control, etc.)
Step 3: Upload the photos of the parts (only about 20 pictures need to be uploaded for each part), and fill in the description information of the parts
Step 4: The merchant staff will share the introduction page with the car buyer, or print the small program code and display it offline
Step 5: The user takes and uploads photos of auto parts, and identifies the parts through similarity image retrieval technology
Step 6: View the details of vehicle parts introduction
