
The core message of this paper is that a simple knowledge distillation approach can effectively transfer the open-world knowledge from a large pre-trained vision-language model to a specialized open-world object detector, achieving better performance for unknown object detection compared to the teacher model.


coremsg

skdf-a-simple-knowledge-distillation-framework-for-distilling-open-vocabulary-knowledge-to-open-world-object-detector


SKDF: A Simple Knowledge Distillation Framework for Distilling Open-Vocabulary Knowledge to Open-World Object Detector


title_rewrite


The core message of this paper is to propose a new YOLO-based open-class incremental object detection framework, YOLOOC, that can effectively discover novel classes during inference while maintaining the performance on previously known classes.


yolooc-a-yolo-based-open-class-incremental-object-detector-with-novel-class-discovery


YOLOOC: A YOLO-based Open-Class Incremental Object Detector with Novel Class Discovery
