Detect all known and unknown objects using many sources and probability calibration with UniDetector
Detect all known and unknown objects using many sources and probability calibration with UniDetector
Detecting Everything in the Open World: Towards Universal Object Detection
arXiv paper abstract https://arxiv.org/abs/2303.11749
arXiv PDF paper https://arxiv.org/pdf/2303.11749.pdf
... address universal object detection, which aims to detect every scene and predict every category ... propose UniDetector, a universal object detector
... UniDetector ... 1) it leverages images of multiple sources and heterogeneous label spaces for training through the alignment of image and text spaces, which guarantees sufficient information for universal representations.
2) it generalizes to the open world easily while keeping the balance between seen and unseen classes, thanks to abundant information from both vision and language modalities.
3) it further promotes the generalization ability to novel categories through ... proposed decoupling training manner and probability calibration.
... UniDetector to detect over 7k categories ... with ... 500 classes participating in training.
... UniDetector behaves the strong zero-shot generalization ... surpasses ... supervised baselines by more than 4% ... state-of-the-art performance with only a 3% amount of training data.
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments