Label-Efficient Online Continual Object Detection in Streaming Video
This work addresses the challenge of continual learning in streaming video with reduced supervision, which is incremental as it builds upon existing continual learners for object detection.
The paper tackles the problem of label-efficient online continual object detection in streaming video by proposing a plug-and-play module, Efficient-CLS, which reduces annotation costs and model retraining time. The method achieves significant improvement with minimal forgetting, outperforming base continual learners with only 25% annotated frames compared to their 100% annotations.
Humans can watch a continuous video stream and effortlessly perform continual acquisition and transfer of new knowledge with minimal supervision yet retaining previously learnt experiences. In contrast, existing continual learning (CL) methods require fully annotated labels to effectively learn from individual frames in a video stream. Here, we examine a more realistic and challenging problem$\unicode{x2014}$Label-Efficient Online Continual Object Detection (LEOCOD) in streaming video. We propose a plug-and-play module, Efficient-CLS, that can be easily inserted into and improve existing continual learners for object detection in video streams with reduced data annotation costs and model retraining time. We show that our method has achieved significant improvement with minimal forgetting across all supervision levels on two challenging CL benchmarks for streaming real-world videos. Remarkably, with only 25% annotated video frames, our method still outperforms the base CL learners, which are trained with 100% annotations on all video frames. The data and source code will be publicly available at https://github.com/showlab/Efficient-CLS.