SUCCESSFUL projects. powered by innovation
Unusual, interesting and exciting projects – that’s what keeps our team going. From healthcare to fashion tech, from robotics to art – with the power of AI we are capable of everything.
features
Self-localization, Object detection, Pose estimation, Navigation, Simulation
client
LG
PROJECT DESCRIPTION
Deep learning Visual Odometry implementation. Fusing measurements from different sensors (IMU, Camera, GPS) using deep learning and Kalman filtering to solve indoor and outdoor localization tasks. Implementation of photorealistic environments and sensor output simulation for synthetic dataset generation and solution testing based on Unreal Engine and Blender.

with reinforcement learning
features
Navigation, Reinforcement Learning, Proximal Policy Optimization
client
LG
PROJECT DESCRIPTION
Coverage Path Problem solving with Reinforcement Learning methods applied to a home robot. We got the solution which is similar to classic algorithms in terms of quality, but which advantage is based on RL methods usage: it is possible for a robot to obtain an adaptive behaviour in extraordinary conditions.

features
Self-localization, Object detection, Pose estimation, Navigation, Simulation
client
NDA
project description
Object position detection in a three-dimensional space using object’s 3D Model and images received from a camera solves the task of Robotic Arm object capture. We used a huge variety of methods for this purpose: Deep Learning, Synthetic Data Generation, classic CV and stereometry methods, AutoML approach. The finalised solution works with several cameras with an up to 10+ FPS Workflow speed.

features
Defect detection on various objects
client
Nienschanz-Automatica
PROJECT DESCRIPTION
For Nienschanz-Automatica we assembled an algorithm that solved the task of detecting defects on various objects. The algorithm is capable of finding an object on an image, receiving information about its textures, detecting texture defects and post-processing of the results.

features
Self-localization, object detection, pose estimation, navigation
technologies
Computer Vision, Artificial Intelligence, Deep Learning
PROJECT DESCRIPTION
OfficeSnail is an autonomous mobile robot that is created to increase productivity of various spaces, including offices, hospitals, factories, etc.
The robot has a number of features, such as autonomous navigation, object and face detection, communication (via voice or an app), task handling, and development of skills that are necessary for a certain task.

features
Pathology analysis; detection and indication of tumors; Improvement of human-powered analysis
client
Intellogic LLC
PROJECT DESCRIPTION
One of our main goals is to make people’s lives easier. That is why working on such project as Botkin.AI is so important for us: with a help from Deep Learning and Computer Vision we are able to provide clinics with better expertise in the field of radiology.
In the course of 2019, we integrated our system to several major clinics across the country, boosting productivity and precision of radiological expertise.

featur
Mobile App for automatic fashion look generation
workflow scheme
B2C
PROJECT DESCRIPTION
Meet OutfitMind — brand-new AI-stylist. OutfitMind is our own app that helps you find the most suitable fashion look for any moment of your life, whether you hang out with your friends, go to the business meeting or just want to walk around on holidays.

features
High Resolution images generation, Poetry generation
technologies
Art Syntesis / Style Transfer
PROJECT DESCRIPTION
While Modern technologies serve as a source of inspiration for us, the Modern technology of our choice — Deep Neural Networks — gets its inspiration from Art in all its forms. Our DNN is an artist, a poet, an artificial creator.
At BrainGarden, we developed Art Synthesising technology based on DNN. All images are generated due to work of an Artificial Intelligence that has all the best Galleries and Museums of the world as its Art School.

client
European University in Saint Petersburg
Features
R&D
PROJECT DESCRIPTION
BrainGarden team conducted research and analysis of modern computer technologies application. The research was also focused on how Artificial Intelligence can be used in art as a medium and in art curation. Our study includes use case scenarios, research technologies, and various artworks, as well as new technologies for their exhibition and the process of artwork creation via machine learning. In the course of our research, we determined a set of effective stable technologies and comprised the map of their application in art.

Does your problem need a solution?
We believe in the power of artificial intelligence.
If you have an idea that you want to realize or if you have any questions please don’t hesitate to contact us – just press the button below.