Object Detection
What's being used in the demo ?
Our model is inspired from YOLO, and is optimized for inference performance on machines without GPU.
Our model achieves state of art mAP 61.2 on the COCO dataset and runs inference 8 times faster than YOLOv3 on CPUs.
Applications
• Can be directly integrated with shopping assistants and shopping bots to introduce easy to use product searching with photos.
• Combined with Image Classification, can be used to robustly analyze
thousands of objects for manufacturing defects, different anomalies.
• With finetuned Object Detection easy to use monitoring systems for various use cases can be deployed.
• Automate shelf object detection, product identification easily with our Object Detection AI.
