Collectibles
Image Tools
Other industries
Our visual AI originated in academia and contributes to the research and development of new technologies today.
Climate change is fundamentally changing the way many sectors of the economy operate, and emphasizes the need for efficiency-enhancing solutions. Satellite monitoring of vegetation, as part of the trend of digitization, significantly increases the efficiency in insurance, agriculture, vegetation management and other areas where it is necessary to know the current vegetation coverage.
However, the unavailability of such data during cloudy periods (spring, autumn) and in cloudy areas (tropical regions, the Baltics, etc.), greatly limits its usability.
That is why World From Space (WFS) and Ximilar are building the LAICA service, which enables daily monitoring and prediction of vegetation status.
The service works by merging multiple types of satellite data using deep learning methods.
The project “LAICA – Satellite monitoring of vegetation development by artificial intelligence methods” (CZ.01.1.02 / 0.0 / 0.0 / 21_374 / 0027239) is co-financed by the European Union under the Operational Program Enterprise and Innovation for Competitiveness.
The aims of the project are to research and develop innovative software that will enable the end-user to effectively use similarity search in the collections of visual data. The key selling point of this software will be the ability to specify a specific type of similarity tailored to customer data and the required application. It is basically an automation of visual search algorithms, which is clearly missing in the market.
The preparation of such a system will be demanding in terms of technology, time, and financing. Three mains outputs are expected: a system for 1) similarity model training automation, 2) management of training data collection, and 3) similarity search operation as a service (SaaS).
E-commerce is rapidly changing with new technologies like AI, Machine Learning (ML), and Computer Vision (CV). Large e-commerce companies have been investing in these areas because they can automate processes and thus save their costs significantly.
However, smaller e-commerce businesses cannot afford the development of such advanced solutions, and they lag behind the big players. In the area of fashion and home decor, smaller businesses often advertise their products via price comparators, because they don’t have the access to the same technologies the larger players have.
Ximilar would like to focus on visual AI services for middle and large e-commerce subjects. Within this project, we would like to make our services accessible to thousands of smaller e-shops that cannot afford the development of their own advanced custom solutions.
The objectives of this project are 1) detailed exploration of major e-commerce platforms and services they provide in the area of image-based machine learning and computer vision, and 2) building a fully functional pilot plugin for a selected service in one selected e-commerce platform.
The objective of the project was to create a platform for collecting audiovisual content from public events such as concerts, festivals, or sports matches, automatically processing it, and producing a video that combined footage from multiple fans.
In the first phase of the project, we focused on IT development, and in the second phase, in collaboration with our partner, the University of Economics, Prague (VŠE), we conducted a series of marketing studies aimed at musical artists, event organizers, and attendees of music events.
Cílem projektu bylo vytvoření platformy pro sběr audio-vizuálního obsahu z hromadných událostí typu koncert, festival nebo sportovní zápas, jejich automatické zpracování a vytvoření videa, které kombinovalo záběry od více fanoušků.
V první části projektu jsme se věnovali samotnému IT vývoji a v druhé části jsme ve spolupráci s partnerem VŠE, Praha vytvořili sérii marketingových studií se zaměřením na hudební umělce, organizátory a návštěvníky hudebních akcí. Spolufinancováno Evropskou unií.