What we produce
Aloception Bird Eye View
The BEV is an enriched representation of a navigable scene. All essential information to navigate through an environment are concentrated into one single output.
Aloception can run on various setups of vision system (monocular-based, stereo-based, multiset). It provides essential and high-value information in real time to enable a more efficient and fluid navigation.
ON-BOARD AI
ON-CAMERA AI
Essential information for navigation
Aloception provides four main generic layers of information for navigation. These information are the main components of the Bird Eye View.
Navigable areas
Obstacle detection
People detection
Localisation & Mapping
Navigable areas
Obstacle detection
People detection
Localisation & Mapping
Same output, different environments
Rain, glare, reflective ice surface, low light/night-time, speed blur effect, and other difficult conditions have been used to train Aloception.
This variety of complex environments has been a baseline to test and ensure the relevance and the robustness of the scene understanding process.
Aloception Development Kit
To discover Aloception by yourself