hustvl/QueryInst
A query-based instance segmentation model published at ICCV 2021, built on top of mmdetection.
Not currently ranked — collecting fresh signals.
star history
QueryInst treats instance segmentation as a query-based detection problem using transformer attention mechanisms. It extends the query-based detection paradigm (similar to DETR) to simultaneously handle object detection and mask segmentation. The model was integrated into the OpenMMLab mmdetection library and achieves competitive performance on COCO benchmark for both instance segmentation and object detection.
Frequently asked
- What is hustvl/QueryInst?
- A query-based instance segmentation model published at ICCV 2021, built on top of mmdetection.
- Is QueryInst open source?
- Yes — hustvl/QueryInst is open source, released under the MIT license.
- What language is QueryInst written in?
- hustvl/QueryInst is primarily written in Python.
- How popular is QueryInst?
- hustvl/QueryInst has 415 stars on GitHub.
- Where can I find QueryInst?
- hustvl/QueryInst is on GitHub at https://github.com/hustvl/QueryInst.