3D head reconstruction solution
Our client is a creative design agency that provides branding, advertising, graphic design, and virtual modeling services for individuals and corporate customers. The agency has its own in-house souvenir production.
Image recognition service
In Japan, many people send their children between the ages of 0 and 5 to preschool. School photography in preschool is a good method of control. It’s a Japanese children's face recognition app designed to make it easier for preschool workers to report on the time spent in preschool.
Project background
In Japan, many people send their children between the ages of 0 and 5 to preschool. Children are there on weekdays from morning until 3-6 pm, while their parents work. Parents want to be aware of how their children spend time at preschool.
The conditions for children at preschool institutions are monitored through regular reporting established by these institutions.
Still, some parents want more than mere formalized reporting. Photos of children at preschool are a good way to confirm a child’s condition. The goal of this project was to help educators to use face recognition technology in providing visual information about a child’s condition.
The biggest challenge was to design and apply neural networks that work on mobile devices with good accuracy and performance.
- Jun 2021 – Sep 2021
- Japan
- Photography
- PyTorch mobile
- Tensorflow-lite-vision
- Tensorflow-lite-meta
- Kotlin-stdlib
- yolov5n
- mobileNetV3_face_land
- moibleFaceNet_m1sm
- Registering a child in the app A teacher can take a photo from the camera or gallery, label a user, save or cancel the labeling process, and see the list of users.
- Processing a photo The photo processing procedure allows a teacher to take a photo from the camera or gallery, see the labeling near the user, take pictures of the user.
- Cancellation A teacher can cancel if there are no required users (children) for reporting.
Business needs
The goal of the project was to help educators to use face recognition technologies in the following:
- Simplify the interaction of educators with the recipients (parents) of the child depicted in the photo.
- Display the previous number of photos and the photos of each child sent previously.
- During the shooting, mark the children whose photos it is forbidden to take.
Solution
CHI Software Development team was involved from the very beginning. The peculiarity of the application is that all the logic of the system was implemented on the mobile device, without a back-end part. We solved the following issues:
-
Registration of children (the ability to take a photo of one child and save it to the database on the device)
-
Processing of the faces in the photo in real-time
-
Comparing the detected face images with the face images in the database and labeling the found faces
-
We had limitations on the accuracy of the F1 measure not lower than 85% and the speed of the system not exceeding 3/4 of a second
-
We conducted research, tried, and tested more than 10 neural networks for obtaining landmarks and face embeddings. Most of them had limitations that didn't allow us to run them locally on the device. But at the end of the day, we succeeded
Client values
- Using facial recognition technologies, the CHI Software team helped preschool workers to produce fast reports for parents.
- In the POC phase, we developed functionality for registering children and searching for faces in real-time.
- We managed to achieve the processing and recognition of 5 faces at the same time and the speed of 3/4 second in real-time.
Employee testimonial
I am a project manager at CHI Software. I took overall project management activities in our project. The project was divided into two phases: a development phase and an optimization phase. We delivered a working app within 4 months. Our team consisted of 3 experienced professionals. We had technologies like Android and Computer Vision at our disposal. We managed to achieve an excellent result and simplify the work of the staff of preschool institutions. With the introduction of our application, the reporting work has been optimized.
Hanna Vinokurova Project Manager