Continual Universal Object Detection
By Xialei Liu, Hao Yang, Avinash Ravichandran, Rahul Bhotika, Stefano Soatto. 2020
13 citations recorded by the source corpus.
Object detection has improved significantly in recent years on multiple challenging benchmarks. However, most existing detectors are still domain-specific, where the models are trained and tested on a single domain. When adapting these detectors to new domains, they often suffer from catastrophic forgetting of previous knowledge. In this paper, we propose a continual object detector that can learn sequentially from different domains without forgetting. First, we explore learning the object detector continually in different scenarios across various domains and categories. Learning from the analysis, we propose attentive feature distillation leveraging both bottom-up and top-down attentions to mitigate forgetting. It takes advantage of attention to ignore the noisy background information and feature distillation to provide strong supervision. Finally, for the most challenging scenarios, we propose an adaptive exemplar sampling method to leverage exemplars from previous tasks for less forgetting effectively. The experimental results show the excellent performance of our proposed method in three different scenarios across seven different object detection datasets.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Tyler L. Hayes
- Christopher Kanan
- Maxim Bazhenov
- Giri Panamoottil Krishnan
- Xuefeng Du
- Mu Cai
- Zhaoning Wang
- Yongtao Wang
In Computer Vision and Pattern Recognition
- Tyler L. Hayes
- Christopher Kanan
- Xuefeng Du
- Maxim Bazhenov
- Giri Panamoottil Krishnan
- Mu Cai
- Zhaoning Wang
- Yixuan Li
Citations in the corpus
- Replay in Deep Learning: Current Approaches and Missing Biological Elements (2021, positive): “In Acharya, Hayes, and Kanan (2020) and Liu, Yang, Ravichandran, Bhotika, and Soatto (2020), replay is used to mitigate forgetting for a continual object detection approach.”
- Continual Contrastive Self-supervised Learning for Image Classification (2021, background): “Recently, most researchers focus on overcoming catastrophic forgetting in the supervised tasks, including image classification[38][31][32][3][28] and object detection[33][44], etc.”
- VOS: Learning What You Don't Know by Virtual Outlier Synthesis (2022, methodology): “Open-world object detection includes out-of-domain generalization (Kim et al., 2021; Wang et al., 2021), zero-shot object detection (Gu et al., 2022; Rahman et al., 2020) and incremental object detection (Liu et al., 2020b; P´erez-R´ua et al., 2020).”
- Comprehensive Attention Self-Distillation for Weakly-Supervised Object Detection (2020, background): “The student networks mimic the the teacher networks on predictive probabilities [32, 34], intermediate features [35], or attention maps of intermediate neural activations [36, 37, 38, 39].”