Custom object recognition. Industry leading.
From the WizyVision Media Library, datasets can be created easily by importing images or reusing existing assets.
Experts have the ability to annotate images, by adding Tags to designate those objects to recognize.
Annotation is done manually in a dedicated interface designed for quick and easy use.
Custom model training
Launch machine learning in a single click. The performance of the model can be visualised after a short period of training. At this point the decision can be made as to whether additional training is needed or the model can be activated.
WizyVision assists in keeping control of ML expenses through its unique option to allocate a budget for training ML.
Custom model deployment
Deploy a custom object recognition model in just one click.
Object recognition can be applied to any imported image or video or to specific ones only.
Human in the loop
Humans can review and correct automatic detections to improve the performance of custom object detection if needed.
WizyVision has 2 levels of human feedback:
- Any user can contribute, and thus feed the expert review level
- Experts can correct annotations performed automatically, optionally by picking images already reported by users.
WizyVision. Become a Machine Learning Expert in 2 clicks.
Allow your experts to annotate images easily in a dedicated, no-code interface.
Anybody can create custom object recognition models. No Machine learning expertise required. At all.
Build a continuous improvement loop to enhance the performance of your custom models in a few clicks. Again, no code required.